Correcting the imbalanced protective RAS in COVID-19 with angiotensin AT2-receptor agonists

The severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) that is responsible for the global corona virus disease 2019 (COVID-19) pandemic enters host cells via a mechanism that includes binding to angiotensin converting enzyme (ACE) 2 (ACE2). Membrane-bound ACE2 is depleted as a result of this entry mechanism. The consequence is that the protective renin–angiotensin system (RAS), of which ACE2 is an essential component, is compromised through lack of production of the protective peptides angiotensin-(1-7) and angiotensin-(1-9), and therefore decreased stimulation of Mas (receptor Mas) and angiotensin AT 2 -receptors (AT 2 Rs), while angiotensin AT 1 -receptors (AT 1 Rs) are overstimulated due to less degradation of angiotensin II (Ang II) by ACE2. The protective RAS has numerous beneficial actions, including anti-inflammatory, anti-coagulative, anti-fibrotic effects along with endothelial and neural protection; opposite to the deleterious effects caused by height-ened stimulation of angiotensin AT 1 R. Given that patients with severe COVID-19 exhibit an excessive immune response, endothelial dysfunction, increased clotting, thromboses and stroke, enhancing the activity of the protective RAS is likely beneficial. In this article, we discuss the evidence for a dysfunctional protective RAS in COVID and develop a rationale that the protective RAS imbalance in COVID-19 may be corrected by using AT 2 R agonists. We further review preclinical studies with AT 2 R agonists which suggest that AT 2 R stimulation may be therapeutically effective to treat COVID-19-induced disorders of various organ systems such as lung, vasculature, or the brain. Finally, we provide information on the design of a clinical trial in which patients with COVID-19 were treated with the AT 2 R agonist Compound 21 (C21). This trial has been completed, but results have not yet been reported.


Introduction
The discovery in 2003 of angiotensin converting enzyme (ACE) 2 (ACE2) as the binding site and cellular entry point for the severe acute respiratory syndrome coronavirus (SARS-CoV) unexpectedly pointed to a link between coronavirus infections and the renin-angiotensin system (RAS) [1]. The current global Corona Virus Disease 2019  pandemic caused by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), which also uses ACE2 as a cellular entry point [2,3], has revived interest and research in this area, resulting in thousands of publications during the past 11 months.
ACE2 is highly expressed in lung alveolar type II and nasal epithelia [4], and therefore these cells serve as main entry points for SARS-CoV-2 into the body. Many of the individuals who contract SARS-CoV-2 are either asymptomatic, or experience mild, common cold-like symptoms and recover soon. However, ∼20% of patients develop more serious COVID-19 disease, which is often driven by an excessive immune response termed as a 'cytokine storm' [5]. In most cases, these patients develop pneumonia, which can lead to respiratory failure and acute respiratory distress syndrome (ARDS) [6]. Many patients also experience extra-pulmonary manifestations affecting heart, kidneys, brain and other organs with multi-organ failure

Interaction of the AT 2 R with other components of the protective RAS
The interactions of the AT 2 R with other components of the protective RAS are multifaceted. As will be discussed in detail in later paragraphs of this review, these interactions are important to understand the rationale of AT 2 R-agonists for the treatment of COVID- 19. In general, according to current data it seems that the different arms of the protective RAS, i.e. those connected to the AT 2 R and those connected to Mas/MrgD, reinforce each other by various ways of cross-talk. The main examples of this type of cross-talk are discussed in the following paragraphs:

Effect of AT 2 R stimulation on the expression of other protective RAS components
There are multiple publications reporting that AT 2 R stimulation leads to an increase in the expression of other protective RAS components thus constituting a type of positive feedback mechanism within the protective RAS: for example, stimulation of AT 2 R leads to an increase in ACE2 expression and subsequently an increase in the protective RAS hormones Ang-(1-7) and Ang- (1)(2)(3)(4)(5)(6)(7)(8)(9). To name some examples, an AT 2 R-mediated increase in ACE2 expression was demonstrated in the renal cortex of male, obese rats on a high-salt diet [23] and of male obese Zucker rats [24], in the plasma of type I diabetic rats with ischaemic renal injury [25], and, importantly, in fibrotic lungs in a monocrotaline model in mice [26].
An AT 2 R-mediated increase in Mas expression was, for example, reported in the renal cortex of obese Zucker rats and in lungs of rats after monocrotaline challenge [24,26].
This positive reinforcement within the protective RAS also works in the other direction with Mas-stimulation resulting in increased AT 2 R expression as, for example, shown in injured femoral arteries of rats treated with Ang-(1-7) [27].

Aligned regulation of protective RAS components
Apart from direct regulation of ACE2 by AT 2 R/Mas stimulation, there is also evidence in the literature that AT 2 R and ACE2 expression is often regulated by various RAS-related or -unrelated compounds in an aligned way. For example, tissue-protective compounds such as AT 1 receptor blockers (ARBs), curcumin [28], glucagon-like peptide-1 receptor agonists or dipeptidyl peptidase-4 inhibitors [29] caused a coincident increase in AT 2 R and ACE2 expression/activity. In contrast, harmful compounds such as Ang II, applied at a hypertensive dose in male Sprague-Dawley rats [30], led to a decrease in both AT 2 R and ACE2 expression/activity. It is tempting to speculate that AT 2 R and ACE2 share a common mechanism of gene regulation, which reacts to certain stimuli in the same way; however, this has not been investigated as yet.

AT 2 R/Mas dimerisation
Another type of AT 2 R-/Mas-axis interaction is the heterodimerisation of AT 2 R and Mas. AT 2 R/Mas-heterodimerisation has first been shown in rat astrocytes in vitro [31] and later in renal proximal tubules and cortical homogenates from obese Zucker rats ex vivo and in human proximal tubule epithelial cells in vitro [32]. Other organs and cell types have not been tested yet for AT 2 R/Mas-heterodimerisation, and consequently knowledge about the occurrence and significance of AT 2 R/Mas-heterodimerisation is still fragmentary. What is known is that at least in some tissues, in which AT 2 R and Mas form heterodimers, functionality of one receptor depends on the presence of the other receptor [31,32].
With regard to cross-talk between the AT 2 R-related and the Mas-related arms of the protective RAS, AT 2 R/Mas-heterodimerisation could cause that AT 2 R-agonists also stimulate Mas and Mas-agonists also stimulate the AT 2 R. Moreover, heterodimerisation seems to cause cross-inhibition, which means that AT 2 R effects can be blocked by a Mas antagonists and Mas effects by an AT 2 R antagonist [31,32].

Effects of Ang-(1-7) and Ang-(1-9) at the AT 2 R
Ang-(1-7) and Ang-(1-9) are both products resulting from ACE2 enzymatic activity with Ang-(1-7) derived from cleavage of Ang II and Ang-(1-9) derived from cleavage of Ang I [22]. In the context of physiological and pathophysiological actions of ACE2, including its role in COVID-19, focus is usually laid on the ACE2/Ang-(1-7)/Mas axis [11]. However, ACE2 enzymatic products are not only agonists for the receptor Mas, but also for the AT 2 R. Specifically, Ang-(1-7) binds to the AT 2 R with low affinity and has agonistic properties at this receptor [33,34], and Ang-(1-9) has been described as an endogenous AT 2 R agonist [35]. However, it is not entirely clear, whether actions of Ang-(1-9) are really elicited by this hormone or whether the active component is actually its degradation product, Ang-(1-7).
The fact that Ang-(1-7) and potentially Ang-(1-9) have agonistic properties at the AT 2 R obviously means that in case of changes in ACE2 expression and enzymatic activity such as in COVID-19, not only signalling and effects mediated by the receptor Mas are affected, but also signalling and effects mediated by the AT 2 R.
In summary, there are several mechanisms, by which an alteration of one RAS component can have an impact on the expression or functionality of other RAS components. The following section will discuss the potential nature of such interactions, specifically how reduced ACE2 expression in COVID-19 may lead to changes within the protective RAS and how this may impact the course of the disease.

Imbalance of the protective RAS in COVID-19
As discovered by Li [2,3], SARS viruses bind to extracellular portions of ACE2 with their spike (S) glycoprotein. The process of SARS-CoVs/ACE2 binding is facilitated through S-protein priming by transmembrane protease serine 2 (TMPRSS2) [36,37]. This protease is a potential drug target for the treatment of COVID-19, since inhibitors of TMPRSS2, camostat mesylate and nafamostat mesylate, successfully prevented SARS-CoV-2 cell entry into cells [37,38].
After binding of SARS-CoV viruses to ACE2, the virus/enzyme complex is internalised, which leads to loss of enzymatically active ACE2 at the cell surface [37,39,40].
The loss of ACE2 enzymes and ACE2 enzymatic activity is the key event that triggers the resulting imbalance within the RAS. There are two main initial consequences of a loss of ACE2 activity: (1) decreased synthesis of the protective mediators Ang-(1-7), Ang-(1-9) and alamandine, and (2) increased levels of Ang II due to less degradation by ACE2 [11]. Or in other words, there is a weakening of the protective arm of the RAS and a strengthening of the potentially harmful, classical arm of the RAS.

Experimental and clinical evidence for an imbalanced RAS in SARS-CoV infection and consequences for the course of disease
As stated above, current knowledge about ACE2 suggests that in case of a SARS-CoV infection and a resulting decrease in ACE2 activity, Ang II levels will rise, while levels of Ang-(1-7) [and also Ang-(1-9)] will decline. In the following, we review studies which tested the above assumption about a RAS imbalance in COVID and whether such an imbalance does have any impact on the course of the disease.
The strongest evidence still comes from studies performed in relation to the first SARS-CoV/COVID outbreak in 2003. A landmark study in this regard is a publication from Josef Penninger's group from 2005. In this study, the authors first demonstrated the crucial role of ACE2 for the process of SARS-CoV infection [39]. They found that ACE2-knockout mice presented with a much lower pulmonary viral load than wild-type mice on day 2 after intranasal inoculation with 100 μl SARS-CoV virus, thus supporting the importance of ACE2 as an entry gate for the SARS-CoV virus. In further experiments within this study, the authors used a recombinant SARS-CoV surface-Spike protein instead of the intact virus after showing that the Spike-Fc protein binds to ACE2 and reduces ACE2 surface expression in vitro. After Spike-Fc protein 'infection' of C57Bl/6 mice, which were additionally challenged by acid aspiration to induce acute lung injury (ALI), the authors indeed reported an imbalance of the RAS by showing that pulmonary Ang II levels, which were already elevated after acid aspiration, increased further by Spike-Fc protein infection [39]. Infection with the Spike-Fc protein and elevated Ang II levels coincided with a more severe course of ALI in these mice. Moreover, the fact that pharmacological AT 1 R blockade with losartan (15 mg/kg) lessened the severity of symptoms of ALI in these mice clearly supported the hypothesis that elevated Ang II levels caused by virus-induced lowering of ACE2 expression and acting via the Ang II/AT 1 R axis play a crucial role for the severity of disease [39].
In another study from the same group, the authors showed in three different, non-viral models of ALI and ARDS (models of ALI induced by acid aspiration, sepsis or endotoxins) that the pathological changes within the lung, such as reduced lung elastance and pulmonary edema formation, were more severe and mortality increased in ACE2-deficient mice (a condition resembling SARS-CoV-induced ACE2 inactivation) compared with wild-type mice [41]. The responsibility of ACE2 deficiency for these differences was further substantiated by rescue experiments, in which application of human recombinant ACE2 significantly attenuated the severity of lung injury and dysfunction in ACE2-knockout mice, an effect that was not seen after application of mutated, non-functional recombinant ACE2 molecules.
This study also provided important evidence that ACE2 deficiency indeed causes an increase in Ang II levels by demonstrating elevated Ang II content in lungs of ACE2-knockout mice when compared with wild-type mice in the acid aspiration model [41]. Moreover, similar to the situation in the study with Spike-Fc protein infection, pharmacological AT 1 R blockade lessened the severity of symptoms of ALI in ACE2-knockout mice, thus supporting the detrimental role of elevated Ang II levels caused by ACE2 deficiency. Interestingly in the context of this article, the authors also concluded that the AT 2 R protects against lung injury; however the only data provided in support were the lack of reversal of severe lung injury in ACE2-KO mice by pharmacological AT 2 R blockade [41].
Taken together, these studies by the Penninger group provided strong support for the hypothesis that elevated Ang II levels as a result of virus-induced down-regulation of ACE2 are an important driver of SARS-CoV induced lung injury. A piece of information that was missing in these studies and would have rounded out the story was a proof for the impairment of the protective arm of the RAS, specifically lower levels of Ang-(1-7) and Ang-(1-9) in lungs of animals infected with the SARS-CoV virus, with the Spike-Fc protein or in ACE2-deficient mice.
It is still not really clear how much the imbalance of the RAS as predicted from theoretical considerations and as observed in animal models actually occurs in humans infected with a SARS-CoV virus, and what the impact of such an imbalance on the course of the disease in humans might be.
A few first studies have been published in which plasma concentrations of RAS components in COVID-19 patients were determined during the current outbreak. These studies elicited conflicting results.
One study by Wu et al. measured plasma Ang II levels in 82 patients with COVID-19 and 12 patients who were severely ill from other diseases, and found a significant increase in Ang II levels in COVID-19 patients, which correlated with the severity of the disease [42]. Measurement of Ang II was likely by enzyme-linked immunosorbent assay (ELISA), but no details about methodology were reported. Liu et al. also reported significantly increased Ang II plasma levels in 12 patients with COVID-19, which correlated positively with viral load and lung injury [43]. In contrast, Henry et al. did not note any difference in plasma Ang II levels between 30 patients with COVID-19 and 12 healthy controls [44]. There were also no differences between hospitalised and discharged patients or between patients requiring intensive care unit support or not. The latter two studies used ELISA for the measurement of plasma Ang II.
A very recent study by Kintscher et al. determined plasma angiotensin peptides by LC-MS/MS [45]. Plasma samples were derived from patients with or without COVID-19 admitted to the emergency room. This study also did not find any significant differences in angiotensin peptide levels or ACE2 activity between patients with or without COVID-19. However, as the authors state themselves, the study had several limitations including a low number of patients (n=8-12 per group) and therefore insufficient statistical power.
It is a limitation of all four studies that they only measured Ang II levels in plasma, which do not necessarily reflect the situation in infected tissues like the upper respiratory tract and the lungs.

Mechanisms of weakening the protective actions of the AT 2 R by SARS-CoV-2 infection
While it is discussed many times in the recent literature on RAS/COVID-19 interactions that a weakening of ACE2 enzymatic activity leads to a decrease in synthesis of Ang-(1-7) resulting in reduced stimulation of the receptor Mas and a weakening of the protective effects through β-arrestin-biased agonism at the AT 1 R, there is much less awareness that decreased ACE2 enzymatic activity also leads to reduced stimulation of the AT 2 R. This is for several reasons, which are illustrated in Figure 1 and are as follows: • less synthesis of the putative AT 2 R-agonist Ang-(1-9), • less synthesis of Ang-(1-7), which is not only a Mas, but also a low-affinity AT 2 R-agonist, • less AT 2 R-stimulation by Ang-(1-7) through cross-activation of Mas-AT 2 R-heterodimers, and • less induction of AT 2 R-expression by Ang-(1-7) acting on the receptor Mas.
It is also important to note that decreased ACE2 enzymatic activity in COVID-19 would lead to increased levels of the ACE2 substrate Ang II, that would conceivably stimulate the protective RAS via activation of AT 2 R. However, the benefits of such AT 2 R activation by Ang II would largely be overridden or offset by its simultaneous activation of the more predominant AT 1 R pathways. Thus, application of a selective AT 2 R agonist (or Mas agonist) would be required to stimulate the protective RAS in COVID-19.

Rationale for AT 2 R agonists in COVID-19
One of many current attempts to find a treatment for patients with COVID-19 is to strengthen the protective arm of the RAS, which is impaired in a multifaceted way by infection of cells with the SARS-CoV-2 virus as reviewed above. Three such approaches, which are human recombinant ACE2 [46], formulations of Ang-1-7 (e.g. TXA127) [47] or an AT 2 R agonist [13] are currently being tested in clinical trials in patients with COVID-19.
The following will discuss the rationale for AT 2 R agonists in COVID-19 including considerations on how AT 2 R-agonists may correct for the above-discussed imbalance within the protective RAS and also on how direct effects of AT 2 R stimulation on certain aspects of COVID-19 pathology may have a therapeutic effect in these patients.

Compensation for impairment of the protective RAS
As discussed in the preceding sections, infection of a cell or organism with a SARS-CoV virus leads to a reduction in ACE2 enzymatic activity, resulting in a lessening of the synthesis of the protective RAS mediators Ang-(1-7) and Ang-(1-9) and, subsequently, decreased stimulation of the receptor Mas and the AT 2 R [11].
While it is likely that in SARS-CoV-2 infections substituting ACE2 with recombinant human ACE2 or compensating for reduced Ang-(1-7) levels by application of recombinant Ang-(1-7) (or derivatives) may be promising strategies to correct the virus-induced impairment of the protective RAS, it is less obvious why AT 2 R-agonists may also be used for this purpose.
As discussed in the section on the imbalance of the protective RAS in SARS-CoV infection, AT 2 R stimulation is impaired by various mechanisms. Collectively, these mechanisms weaken AT 2 R stimulation and signalling and this can be counteracted by application of AT 2 R-agonists. Moreover, AT 2 R-agonists may also lead to stimulation of Mas signalling through cross-activation of AT 2 R-Mas-heterodimers [31,48].

Anti-inflammatory effects of AT 2 R activation
As would be expected, COVID-19 infection elicits activation of the innate and adaptive immune systems, and in a majority of cases their combined actions result in disease resolution [49][50][51][52][53]. In patients with more severe COVID-19 disease, those displaying ARDS and pneumonia, clinical findings indicate that multiple aspects of the innate and adaptive immune systems are compromised or dysregulated [50,51,53]. Of importance for the current article is the observation that this group of COVID-19 patients exhibits massively increased levels of circulating pro-inflammatory cytokines, including interleukin (IL)-1β (IL-1β), IL-6, tumour necrosis factor (TNFα), IL-17 and interferon-γ, (IFN-γ) [53,54]. The resulting chronic pro-inflammatory milieu, termed as a 'cytokine storm' , is likely responsible for the uncontrolled inflammation within the lungs, heart, vascular endothelium and kidney, and recruitment of immune cells (macrophages, neutrophils, T cells) to these sites [7,55]. The subsequent release of pro-inflammatory cytokines from these cells and generation of reactive oxygen species results in endothelial cell and tissue damage [8,51,56].
Considering the excessive pro-inflammatory response and the decreased anti-inflammatory capacity of the ACE2/Ang-(1-7) system following COVID-19 infection, anti-inflammatory agents would provide a logical therapeutic strategy [52,70]. Indeed, a number of recent clinical studies have indicated beneficial effects of corticosteroids administered systemically to patients with ARDS in severe COVID-19 disease [71].
It is well-known that AT 2 R agonists exert powerful anti-inflammatory actions [72][73][74][75]; thus, they may be prime candidates to compensate for the loss of ACE2/Ang-(1-7) protective mechanisms and help offset the large pro-inflammatory response in COVID-19 patients with severe disease (Figure 2). Based on the literature available, the anti-inflammatory effects associated with AT 2 R may result from direct effects on immune system cells, as well as effects on non-immune cells.

Innate immune system
In general, initial COVID-19 infection will result in activation of the innate immune system, including monocytes, macrophages, dendritic cells and the complement system [50,51,53]. Human monocytes, macrophages and dendritic cells contain AT 2 R [75][76][77] and a number of studies indicate that activation of this angiotensin receptor subtype interferes with certain aspects of the innate immune system response. For example, in vitro studies have demonstrated that the AT 2 R agonist CGP42112 inhibits the IL-1β induced activation of monocytes [78]. Furthermore, a number of studies indicate that AT 2 R activation interferes with Toll-Like Receptor (TLR) -induced pro-inflammatory mechanisms. TLR-induced activation of macrophages is well established [79], and in the case of COVID-19 infection this involves endosomal TLR3 and TLR7/8, as well as the indirect activation of cell-surface TLR4, via virus-induced oxidised phospholipids [80]. Interestingly, the AT 2 R agonist C21 exhibits anti-inflammatory actions in TLR4-mediated inflammation in THP-1 macrophages, mediated via IL-10 production [76]. Another study, using human monocyte cell lines, has demonstrated that inflammatory responses mediated by TLR4 were attenuated by C21 acting to reduce pro-inflammatory cytokines [81]. Of direct importance to the immune pro-inflammatory response occurring after COVID-19 infection, C21 has been shown to exert significant anti-inflammatory effects in the lung, via influences on TLR4 and macrophage infiltration [82]. Specifically, systemic administration of C21 into rats that had undergone bleomycin-induced pulmonary inflammation and fibrosis prevented the up-regulation of TLR4 and lessened the migration of CD68 + macrophages in the lung [82]. In another study, investigators used neonatal rats with hyperoxia-induced lung injury to evaluate the anti-inflammatory effects of an AT 2 R agonist. Daily treatment of these rats with dKcAng- (1-7), a specific AT 2 R ligand, attenuated the pulmonary influx of macrophages [83]. The anti-inflammatory effects of AT 2 R activation are not confined to monocyte/macrophage-mediated mechanisms, as indirect evidence suggests that dendritic cells are another locus of action. Specifically, human dendritic cells that underwent differentiation in the presence of the AT 2 R antagonist PD123319, displayed a more pro-inflammatory phenotype [84]. A more recent study indicates that the potent anti-inflammatory effect of ACE inhibitors on dendritic cells post-myocardial infarction is at least partly due to activation of AT 2 R [85]. As yet, there is no evidence for modulation of the complement system by AT 2 R agonists.

Adaptive immune system
COVID-19 infection also results in activation of the adaptive immune system, albeit over a more delayed time frame [50,51,53]. T helper (CD4 + ) and cytotoxic (CD8 + ) T cells are central to the adaptive immune response [86]. Interestingly, a feature of COVID-19 disease is that patients display lymphopaenia, with a reduction in the absolute levels of all types of T cells, of natural killer (NK) cells and also B cells [51,53]. While this lymphopaenia would compromise the adaptive immune response, studies have also demonstrated that the activity of CD8 + cells and of pro-inflammatory Th17 cells is increased in COVID-19 patients, and may be partly responsible for tissue damage in patients with severe disease [51,52]. Thus, agents that can counter the adaptive immune system pro-inflammatory response may be of value to COVID-19 patients. A number of studies have concluded that AT 2 R are not expressed at high levels on cells of the adaptive immune system [87,88]. However, T cells and NK cells do express AT 2 R at low levels [89,90], and increased levels of AT 2 R-expressing T cells of CD4 + and CD8 + varieties appear to become important under certain disease conditions with a strong pro-inflammatory component. For example, CD4 + AT 2 R + FoxP3 + were identified as a novel T-regulatory cell subset that was up-regulated in a rat model of myocardial infarction (MI) and in human heart failure patients, and acted anti-inflammatory via the secretion of IL-10 [91]. Myocardial transplantation of these CD4 + AT 2 R + FoxP3 + T cells decreased infarct size and improved cardiac function in the rat MI model [91]. A similar study identified a cardioprotective CD8 + AT2R + T-cell population, which increased during ischaemic heart injury. These cells secreted IL-10 upon AT 2 R stimulation by C21, exhibited decreased IL-2 and IFN-γ expression, and contributed to maintaining viability of cardiomyocytes [92]. The protective-and anti-inflammatory effects of AT 2 R-expressing CD4 + T cells were also apparent in thoracic aortic aneurysm (TAA) [93]. Further studies which utilised naïve T cells isolated from mouse spleen and lymph nodes demonstrated that AT 2 R stimulation with C21 under polarising conditions modified their differentiation, resulting in a less pronounced pro-inflammatory phenotype as indicated by lowered expression of IFN-γ and IL-17 and increased expression of the marker of regulatory T cells, FoxP3 [94]. Despite these studies that demonstrate protective, anti-inflammatory actions of AT 2 R-expressing T cells, one study indicates the opposite. Namely, Caillon et al. [95] demonstrated that AT 2 R-activation of Th17 cells induced secretion of the pro-inflammatory cytokine IL-17 that drove flow-mediated outward arterial remodelling, and proposed that this is necessary for collateral artery growth in ischaemic disease -once again, a beneficial action of AT 2 R activation, but opposite in terms of inflammatory response. As stated above, AT 2 R are also expressed on NK cells (particularly uterine NK cells) [89,96]. However, their specific roles in terms of influencing the immune system are unknown.
Thus, there is evidence that points to direct beneficial actions of AT 2 R activation via cells of the innate and adaptive immune systems, which could be beneficial during the hyperinflammatory state in COVID-19. However, based on the available evidence the greater effect of AT 2 R on the immune system is via actions at non-immune tissues and cells.

Anti-inflammatory AT 2 R actions via non-immune cells and tissues
There is a host of data indicating powerful anti-inflammatory actions of AT 2 R within the parenchymal tissues of multiple organ systems throughout the body [73,75]. This includes the lungs [26,82,83,97], heart [85,92,98,99], kidney [100][101][102][103], vasculature [72,73,102,[104][105][106][107][108] and the central nervous system (CNS) [109][110][111][112][113][114][115][116]. These anti-inflammatory actions of AT 2 R activation at non-immune tissues and cells largely involve interruption of intracellular signalling pathways that lead to production of pro-inflammatory cytokines, and also through induction of pathways that elicit production of anti-inflammatory cytokines such as IL-10 [73,75,116]. The cellular and intracellular actions of AT 2 R have been recently reviewed in detail elsewhere [75,117], and the objective of the current article is not to re-review them here. Rather, we have utilised a diagram (Figure 3) to illustrate and summarise the anti-inflammatory actions of AT 2 R at various tissues/cells, and have included the intracellular signalling pathways that are likely of importance in these actions. Here, we have emphasised those anti-inflammatory effects that would likely be important for helping to counter the deleterious overactivation of pro-inflammatory mechanisms in COVID-19 patients, in particular those that subsequently lead to tissue fibrosis. Also included in the diagram are vascular and CNS actions of AT 2 R, that are discussed in more detail in subsequent sections of this review.
As shown in Figure 3, literature evidence points to several intracellular steps that lead to a reduction in the production of pro-inflammatory cytokines following AT 2 R stimulation [117]. In brief, agonist-induced AT 2 R stimulation results in activation of tyrosine and serine/threonine phosphatases via an inhibitory G-protein [118][119][120]. In turn, these phosphatases make less nuclear factor-κB (NF-κB) available for translocation to the nucleus, and inhibit extracellular signal-regulated kinase 1/2 (Erk 1/2) and Janus kinase (Jak)/signal transducer and activator of transcription (Stat) pathways through dephosphorylation events [72,120]. The ultimate result is less production of pro-inflammatory cytokines. On the other hand, AT 2 R stimulation of tyrosine and serine/threonine phosphatases and of Rac-α serine-threonine kinase (Akt) elicits activation of endothelial nitric oxide synthase (eNOS) and generation of nitric oxide (NO) and cyclic guanosine monophosphate (cGMP), the latter exerting anti-inflammatory effects [121]. While the mechanisms of AT 2 R-induced IL-10 production are not established, it has been demonstrated that they require production of NO [102].
In summary, it is clear that activation of AT 2 R exerts potent anti-inflammatory effects, via direct actions on immune system cells and non-immune tissues. Considering the dysregulated immune system and cytokine-storm observed in COVID-19 disease, selective AT 2 R agonists may represent a viable approach for alleviating this excessive pro-inflammatory situation in COVID-19 patients, and in doing so help to reduce the development of downstream heart, kidney and lung complications (Figure 2).

Endothelial protection and anti-coagulation
Quite early during the pandemic it became clear that COVID-19 patients die not only from viral pneumonia causing ARDS and hypoxic respiratory failure, but also from cardiovascular complications, thrombotic events and multi-organ failure [122,123]. A unifying cause for these quite different pathologies may originate from the vascular endothelium [7][8][9]124].
Generally, there are two potential causes for endotheliopathy in COVID-19: either direct viral infection of endothelial cells or the so-called cytokine storm as described earlier in this review. Evidence indeed exists for the presence of both pathomechanisms in COVID-19-associated endotheliitis. Direct viral infection was demonstrated by Varga et al. and Ackermann et al. in endothelial cells from various vascular beds in biopsies from patients with a severe course of COVID-19 [10,125]. Regarding endothelial cell injury during COVID-19-induced cytokine storm, it is generally known, that increased plasma levels of cytokines cause direct cell injury by binding to their respective membrane receptors and initiating pro-inflammatory signalling cascades and eventually cell apoptosis [55,126,127]. Moreover, cytokines promote recruitment of cytotoxic leucocytes by increasing expression of adhesion molecules on the surface of endothelial cells [128].
Regardless of the origin of the pathological changes in endothelial cells in COVID-19, injury of these cells leads to their dysfunction, which is characterised by a shift of the vascular equilibrium towards more vasoconstriction, loss of the anti-coagulant state, endothelial cell death and barrier-breakdown with subsequent tissue oedema and facilitated infiltration of inflammatory cells [8,9]. In addition, inhibition of ACE2 enzymatic activity in COVID-19 prevents degradation of bradykinin by ACE2 thus promoting bradykinin-mediated vascular leakage [129]. These changes contribute to the development of ARDS, to impaired function of other organs such as kidney, heart or brain and to thromboembolic complications [8,9].
Several clinical studies have indeed reported an increase in markers of endothelial dysfunction and coagulation indices [e.g. von Willebrand factor (VWF) antigen, VWF activity, factor VIII activity, D-dimers, fibrinogen, fibrin degradation product; prolonged prothrombin time and partial thromboplastin time], in patients with COVID-19 [7,122,130]. Importantly, several of these markers correlated positively with the prognosis for the individual patient [130]. Other studies reported microangiopathy, capillary microthrombi and disseminated intravascular coagulation by histological examination of post-mortem lungs from COVID-19 patients [125,131].
There is good experimental evidence to support the assumption that AT 2 R-stimulation will act in a protective manner in the event of SARS-CoV2 induced endothelial injury, although data from preclinical or clinical studies in SARS-CoV2-infected individuals is yet to be obtained. The clinical trial with the AT 2 R-agonist C21 in patients with COVID-19 (reviewed later in this article) will presumably provide such data [13].
Generally, AT 2 R-stimulation promotes mechanisms which are beneficial for endothelial function such as activation of eNOS leading to NO generation or synthesis of the vasodilatory mediator epoxyeicosatrienoic acid from arachidonic acid [121,132,133].
Additionally -as reviewed above -stimulation of the AT 2 R acts anti-inflammatory, which also applies to the vasculature. A study by Sampson et al. in human umbilical vein endothelial cells and in ApoE −/− mice fed a high-fat diet provided an in-depth examination of potential anti-inflammatory effects of AT 2 R-stimulation by C21 in the vasculature and found an inhibition of adhesion molecule expression and monocyte adhesion in vitro and an attenuation of leucocyte adhesion in vivo [105]. The authors could further show an AT 2 R-mediated inhibition of NF-κB activity resulting in reduced IL-6 and TNFα cytokine expression. This latter observation confirmed prior reports about an inhibition of the key regulator of cytokine transcription, NF-κB, by AT 2 R-stimulation in rat fetal vascular smooth muscle cells and human dermal fibroblasts [72,134]. An anti-inflammatory effect in vasculature, i.e. a reduction in monocyte infiltration in aortas from deoxycorticosterone acetate-hypertensive rats, has also been shown for the ACE2-dependent, endogenous AT 2 R-agonist Ang-1-9 [135]. Notably, both studies identified AT 2 R-mediated, anti-inflammatory effects, which specifically counteracted pro-inflammatory mechanisms relevant for COVID-19-associated endothelial injury.
As stated above, endothelial barrier breakdown and increased vascular permeability are important factors contributing to tissue oedema, which is a main reason for impaired gas exchange in ARDS [136]. By measuring hydraulic permeability in rat mesenteric venules, Ereso et al. could show that AT 2 R-stimulation by the selective agonist CGP42112A or by Ang II under concomitant AT 1 R-blockade attenuated microvascular fluid leakage, which had been induced by platelet activating factor [137]. Further evidence for a reduction in vascular leakage by AT 2 R-stimulation comes from a stroke model in mice, in which treatment with C21 decreased blood-brain barrier (BBB) permeability and subsequent development of cerebral oedema [111].
Finally, a few studies have suggested that the AT 2 R has anti-thrombotic effects, which could help to counteract the pro-coagulant state in COVID-19. Chabielska et al. described that the AT 1 R-blocker Losartan significantly reduced thrombus weight in a model of vena cava thrombosis in Wistar rats with two-kidney, one-clip induced renal hypertension [138]. The effect of Losartan was blocked by the AT 2 R-antagonist PD123319 indicating that indirect AT 2 R-stimulation by elevated levels of Ang II acting on the unopposed AT 2 R was the underlying mechanism of the Losartan effect.
Interestingly in the context of COVID-19 induced hypercoagulability, a study by Balia et al. found an inhibitory effect of AT 2 R-stimulation on LPS-induced tissue factor expression in peripheral blood mononuclear cells [139]. Apart from mononuclear cells, tissue factor is also highly expressed by vascular, subendothelial cells and is essential for the initiation of the clotting cascade in response to inflammation or tissue injury [140]. Tissue factor on monocytes, which is increased by elevated cytokine levels and also by Ang II, plays a key role in triggering disseminated intravascular coagulation and thrombotic microangiopathy in bacterial or viral sepsis [140]. Increased tissue factor expression on monocytes has also been shown for patients with a severe course of COVID-19, and expression levels correlated with severity of disease [141]. Consequently, experimental data by Balia et al. showing an AT 2 R-mediated reduction in tissue factor expression on monocytes points to a potential therapeutic effect of AT 2 R-agonists in COVID-19 related coagulopathy, which, however, still needs to be proven in clinical studies [139].

Anti-fibrotic effects of AT2R activation
It is established that pulmonary inflammation is a major causative factor in the development of pulmonary fibrosis [142,143]. As indicated in Figure 2 one of the major long-term outcomes in COVID-19 is the development of pulmonary fibrosis [66][67][68][69], likely due to the hyperinflammatory state. In fact, pulmonary fibrosis is one of the pathological situations in which AT 2 R agonists may have beneficial effects for COVID-19 patients. This assumption is based on several preclinical studies in which AT 2 R-agonists were proven to have profound anti-fibrotic activity within many disease states, including pulmonary-, cardiac-, renal-, aortic and pancreatic fibroses. These anti-fibrotic actions have recently been extensively reviewed [117,144], and so the description of all of these actions will not be recapitulated here. Rather, we focus on the pulmonary anti-fibrotic actions, that are likely of direct importance to patients with COVID-19 disease.
Two preclinical studies have demonstrated that the AT 2 R agonist C21 prevents and reverses pulmonary fibrosis. In the first, daily systemic treatment of rats with C21 beginning 2 weeks after induction of pulmonary fibrosis with monocrotaline reversed interstitial and perivascular fibrosis [26]. In addition, C21 treatment decreased transforming growth factor-β mRNA expression, which was greatly increased in the monocrotaline-treated rats [26]. In the second study from the same group, pulmonary fibrosis was induced in rats by intra-tracheal administration of bleomycin and systemic C21 administration began either at the same time as bleomycin (prevention) or 3 days later (treatment); in both cases, C21 administration continued on a daily basis for 2 weeks. In the prevention paradigm, C21 significantly attenuated the formation of pulmonary fibrosis and in the treatment paradigm, fibrosis progression was almost completely stopped from the start of the day of the treatment [82]. Associated markers of fibrosis, such as collagens 1 and 3, connective tissue growth factor, IL-13, and tissue inhibitor of matrix-metalloproteinases were all reduced by C21-treatment [82].
The importance of these findings is manifold: they not only proved the principle that AT 2 R activation by an agonist is beneficial in pulmonary fibrosis, they also laid the groundwork for the development of C21 as a potential therapeutic for idiopathic pulmonary fibrosis [145]. In addition to the above described anti-fibrotic actions, the attenuating effects of AT 2 R agonists on pulmonary inflammation will presumably have a preventive effect with regard to the development of pulmonary fibrosis in patients with a severe, hyperinflammatory courses of COVID-19 [26,82,97]. However, this assumption will have to be proven in future clinical studies. The ongoing trial with C21 in patients with COVID is most likely too short (7 days treatment) and does not include enough patients to detect any potential effects of C21 on the development of pulmonary fibrosis [13].

CNS effects of AT 2 R activation
The most common neurological complications of severe COVID-19 are encephalitis and ischaemic stroke [146,147].
Ischaemic stroke associated with COVID-19 disease occurs in patients of all ages, with more severe strokes, worse functional outcomes and higher mortality rates than are seen in a comparable group of non-COVID-19 patients [148][149][150][151][152]. Thus far, it appears that the major causes for these ischaemic strokes are derived from the hyperinflammatory state and resulting coagulopathy observed in COVID-19 patients, which lead to thromboembolism [153], neurovascular endothelial dysfunction, BBB breakdown, and consequent generalised activation of the CNS innate immune system [149]. Cardiomyopathy, and consequent intracardiac formation of thrombi, which may travel to the brain, is another potential cause of ischaemic stroke in COVID-19 [153].
AT 2 R agonists appear well-positioned to be of significant benefit to COVID-19 patients who have developed ischaemic strokes. Preclinical studies in rats and mice from multiple independent investigative teams, using several different models of ischaemic stroke, have demonstrated the efficacy of AT 2 R agonists applied post-stroke in reducing the severity of intracerebral infarcts, and improving behavioural/neurological outcomes [110][111][112]115,[154][155][156][157][158]. Details of these studies and the mechanisms of neuroprotection have been reviewed by us, recently [159]. The factor that might make AT 2 R agonists a viable approach for ischaemic stroke in COVID-19 patients is their multiple modes of action against stroke that include neuroprotective, vascular and regenerative aspects [159]. The neuroprotective aspects include anti-inflammatory, anti-apoptotic and anti-oxidant effects [159]. Thus, in COVID-19 induced strokes AT 2 R agonist-induced anti-inflammatory and anti-oxidant effects might not only help to depress the cytokine storm, alleviate neurovascular endothelial dysfunction and coagulation, but also suppress the generalised activation of the innate CNS immune system. A further mode of action of AT 2 R agonists in ischaemic stroke is restoring BBB integrity -a clear benefit in severe pro-inflammatory conditions where the BBB is compromised [111]. A final important aspect is that our group has demonstrated that an AT 2 R agonist (C21) can be applied directly to the brain via an intranasal (nose to brain) route, where it exerted significant protective effects against ischaemic stroke in a rodent model [157]. Such a delivery system can easily be replicated in humans.
To summarise, based upon preclinical findings there is strong potential that AT 2 R agonists may be effective in at least alleviating the severity of ischaemic strokes in COVID-19 patients. The caveat is that thus far, there has been no clinical evaluation of the actions of AT 2 R agonists in human stroke patients.

Clinical trial of C21 in patients with COVID-19
The non-peptide, orally active AT 2 R-agonist C21 (Vicore Pharma, Gothenburg, Sweden; now also termed as VP01) has been tested for safety and efficacy in a Phase II clinical trial in patients with COVID-19, the so called ATTRACT In the opinion of the Investigator, the subject will be able to comply with the requirements of the protocol Female subjects of childbearing potential not willing to use contraceptive methods as described in detail in the study protocol Male subjects not willing to use contraceptive methods as described in detail in the study protocol Subjects known or suspected of not being able to comply with this trial protocol (e.g. due to alcoholism, drug dependency or psychological disorder) trial (https://www.clinicaltrials.gov/ct2/show/NCT04452435 and https://vicorepharma.com/the-attract-study/) [13]. The trial was completed on 13 October 2020 (final data collection date for primary outcome measure), but has not yet reported any results (status on 13 November 2020). ATTRACT is a randomised, triple-blind, placebo-controlled trial with parallel design, for which 106 patients were enrolled and randomised to receive either placebo or C21 (100 mg twice daily) po. Treatment duration was 7 days, and C21 or placebo were added on top of standard of care medication. With regard to inclusion and exclusion criteria, it is important to note that only a specific subgroup of patients diagnosed for COVID-19 was eligible for recruitment, which were patients hospitalised for COVID-19 associated, respiratory symptoms, but not requiring intensive care or mechanical invasive or non-invasive ventilation. This specific subgroup of patients was most likely selected for targeting patients with a developing or already existing hyperinflammatory response, but not yet suffering from life-threatening complications. Primary outcome was defined as the change in C-reactive protein (CRP) from treatment start to the end of the 7-day treatment period, i.e. the trial has mainly tested, whether AT 2 R-stimulation is able to attenuate the inflammatory reaction in COVID-19 and prevent or milden the cytokine storm. More detailed inclusion and exclusion criteria can be found in Table 1. The study was performed as a multi-centre trial located in the U.K. and India. ATTRACT is expected to report in late 2020.

Conclusions
The COVID-19 pandemic that is a result of infection with the SARS-CoV-2 coronavirus is affecting millions of people and the death toll has now exceeded 1.3 million victims. It is one of the major challenges in medical research of our time and one of the biggest unmet medical needs. In an enormous effort, scientists all over the world are trying to develop vaccines and effective treatments to bring down the number of COVID-19 associated fatalities and severe courses of disease and to get us all back to a life with normal social interactions and daily routines. One approach in the search for a treatment of COVID-19 is to find drugs that counteract the imbalance of the RAS in this disease, which is caused by loss of enzymatic activity of ACE2, the enzyme which serves as the binding site for