Angiotensin converting enzyme II (ACE2), a type I transmembrane mono-carboxypeptidase of renin angiotensin system (RAS), in involved in conversion of angiotensin I (Ang I) and angiotensin II (Ang II) to angiotensin (1-9) and angiotensin (1-7), respectively. This enzyme, as the receptor of SARS-CoV-2, plays a crucial role in the virus entrance into the host cells. The docking of the S protein to this receptor, eventually leads to the fusion of the virus membrane with the host cell plasma membrane to release the viral genome into the cell cytoplasm.
S protein of SARS-CoV-2 (of 1273 amino acid residues), a class I fusion and homo-trimeric glycoprotein, is composed of two subunits, S1 and S2, in each of its monomers. The former is in charge of docking of the virion to ACE2 and the latter promotes fusion of viral and host cell lipid outer membranes through a fusogenic process [1]. From N- toward C-terminal ends, S1 comprises a signaling sequence, an N-terminal domain (NTD), a receptor binding domain (RBD) including its receptor binding motif (RBM) and two subdomains (SD1 and SD2) [2][3]. S2, a rather conserved molecule in SARS-CoV and SARS-CoV-2, consists of a fusion peptide (FP), the heptad repeat 1 (HR1), the heptad repeat 2 (HR2), a transmembrane domain (TM) and a cytoplasmic tail (CT) [1][4]. Absent in SARS-CoV, a polybasic amino acid sequence, RRAR (Arg-Arg-Ala-Arg), called the furin-cleavage site is located at the S1/S2 boundary, the deletion of which attenuates SARS-CoV-2 replication in the respiratory cells [5]. (Figure 1). It has been shown that the presence of the RRAR motif enables SARS-CoV-2 to evade endosomal interferon-induced transmembrane (IFITM) proteins [6].
Figure 1. S protein (monomer). PDB (RCSB.org) ID: 7QUS [7]. NTD: N-terminal Domain, RBD: Receptor Binding Domain
ACE2 contributes to infectivity and the entry of SARS-CoV-2 into the host cells while physiologically plays a crucial role in protection of the lungs, heart and other tissues against a variety of inflammatory or hypoxia-induced insults [20][21][22][23]. The non-homogenous distribution of ACE2, yet in a limited scale compared to ACE, was shown in a wide variety of tissues with the highest expression in the small intestine, heart, testis, thyroid glands and adipose tissue; intermediate expression in the lungs, colon, liver, bladder and adrenal glands; and the lowest content in the blood, spleen, bone marrow, brain, blood vessels and muscles [24]. Beyond the genetic traits determined by loci near immune related genes such as IFNAR2 and CXCR6 [25], the distinct genetic variants of highly polymorphic Ace2 (gene located on chromosome Xp22.2 encoding for ACE2 protein [26]) in different populations have also been shown to affect the susceptibility to and severity of COVID-19 in some cohorts of patients [27][28][29][30].
ACE2, a type I transmembrane protein of 805 residues, comprises an extracellular segment (residues 19–740) including a signal peptide of 18 amino acids, an N-terminal claw-like protease domain (PDACE2, residues 19–615) and a collectrin-like domain (residues 616–740) including ferredoxin-like fold “Neck” domain (residues 616–726) attached to a long transmembrane domain (residues 741–763) ending with a cytosolic C-terminal tail [31][32][33][34][35]. The cytosolic tail containing a conserved endocytosis short linear motif had been demonstrated to play a role in SARS-CoV-S protein-induced shedding of ACE2, TNF-α production and SARS-CoV infection [36][37] although a combined in silico and in vitro study strengthened by confocal imaging denied any role for C-terminal tail of ACE2 in SARS-CoV and SARS-CoV-2 cell entry [38]. The collectrin-like domain contributes to dimerization of two ACE2 molecules (assuming ACE2-A and ACE2-B) through interacting with Arg652, Glu653, Ser709, Arg710 and Asp713 in ACE2-A with Tyr641, Tyr633, Asn638, Glu639, Asn636 and Arg716 in ACE2-B [31].
Regulated selective cleavage and release of the extracellular segment of many of transmembrane proteins into the extracellular space called “ectodomain shedding” modifies a diverse array of trans- and cis-signaling pathways [52]. ADAM17, the first identified shedding protease, is a Zn2+-dependent metalloproteinase and a member of the family of membrane-anchored ADAMs, which plays a role in both innate and humoral adaptive immunity, among the others [53]. This sheddase promotes proinflammatory effects through dissociating a variety of ligand proteins including ACE2, TNF-α and transmembrane CX3CL as its substrate as well as transmembrane receptors including IL6Ra and TNF receptors I and II [47][54][55][56]. This metalloprotease exists in two forms: the full length pro-ADAM17 of 100KDa and the mature form of 80KDa lacking the inhibitory prodomain: the latter comprises almost two thirds of its total cell content which resides predominantly in perinuclear space along with TNF-α [57][58]. ADAM17 as a type I transmembrane multidomain proteinase comprises an N-terminal signal sequence, an inhibitory prodomain, a metalloproteinase catalytic domain, a disintegrin-like domain, a cysteine-rich and membrane proximal domains (MPD, involved in substrate recognition) attached to a single transmembrane domain ending into a cytoplasmic tail [59][60]. The transmembrane domain, but not the cytoplasmic domain, contributes to the rapid activation of this metalloprotease by different signaling pathways [61]. However, phosphorylation of the cytoplasmic domain of ADAM17 by mitogen activating protein kinase (MAPK) network including p38MAPK and extracellular signal-regulated kinase (ERK) as well as Polo-like kinase2 (PLK2) keep this metalloprotease proteolytically functional through prevention of its dimerization and dampening of its binding to tissue inhibitor of metalloproteinase3 (TIMP3) at the cell surface [62][63]; TIMP3 plays an inhibitory role for the dimerized ADAM17 on the cell surface [64].
Given that ADAM17 small interfering RNA (siRNA) is capable of attenuating Ang II-mediated inflammation in vascular smooth muscle cells [70] it is implied that along with other downstream mediators, ADAM17 is also involved in Ang II-induced inflammatory responses. Additionally, Ang II-mediated stimulation of AT1R, a G-protein coupled receptor (GPCR), provokes oxidative stress, induces phospholipase C (PLC)-mediated protein kinase C (PKC) activation and increases calcium influx [71][72][73][74]. Oxidative stress in tumor cells and platelets could previously be found to activate ADAM17 with pro-inflammatory effects (see previous paragraph) [75][76]. Reciprocally, ADAM17 in a mouse model could also increase NADPH oxidase 4 (Nox4) activity resulting in oxidative stress [77] which by inducing pro-inflammatory genes is intertwined with inflammation [78][79]. It is, however, noticeable that ADAM17 positively regulates Thioredoxin-1 (Trx-1) activity as the key effector of intracellular reducing system and downregulates ADAM17 as a negative feedback effect [80][81].
Furthermore, the PKC pathway, depending on the nature of its activator, affects ADAM17-dependent ectodomain shedding. Short-term (minutes) activation of protein kinase C (PKC) by phorbol ester (the strongest non-physiologic stimulator of PKC) increases ADAM17 content on the cell plasma membrane and the sheddase activity while prolonged (hours) exposure to phorbol ester downregulates mature form of ADAM17 on the cell surface abolishing the ectodomain shedding without reducing the total cell content of ADAM17 (mature plus pro-ADAM17)[82]. Conversely, ligand-activated GPCRs, such as thrombin-mediated protease activated receptor1 (PAR1), with the potential to activating of PKC signaling [83][84], induces sheddase function without any change in the content of ADAM17 on the plasma membrane [82]. Moreover, Ang II, via AT1R, was found to upregulate PAR1 and PAR2 in rat aorta which lead to pro-inflammatory responses accompanied by raising IL-6 and monocyte chemoattractant protein-1 (MCP-1), the substrates of ADAM17 [58][85][86]. Along with the natural homo- and hetero-merization of GPCRs [72], an in vitro study revealed synergistic interaction between AT1R and PAR1 [87], implicating a positive influence of these receptors in activating ADAM17. Moreover, AT1R stimulation alone is associated with a rise in IL-1β which by itself promotes ectodomain shedding by ADAM17 [88][89]. Considering that Ang II via AT1R could induce PKCδ/p38MAPK [90] which also activates ADAM17 [91][92], it is logically implied that AT1R also potentiates ADAM17-induced ectodomain shedding through p38MAPK pathway. Ca2+ influx and calmodulin inhibition stimulate ADAM10 [93]. Given that ADAM17 is involved in processing of pro-α2δ-1 and α2δ-3 subunits of voltage gated calcium channels which enhances calcium influx [94] it is implied that ADAM17 also indirectly contributes to activating ADAM10 in ectodomain shedding.
As was mentioned previously, ADAM10 and ADAM17, the predominant sheddases, contribute to cleaving mACE2 off the cell surface which lead to a vast array of physiological and pathological cis- and trans-signaling [54]. Cleavage of the ectodomain of mACE2 occurs through a basal low level constitutive or a metalloproteinase-dependent fashion [95]. A cell culture study showed that inhibition of both ADAM17 and ADAM10, widely found on pneumocyte type I and II, results in ACE2 increment on the surface of these cells [96]. Consistently, an animal model of diabetes proved that ADAM17 gene knockdown and overexpression could reduce and raise mACE2 shedding off the cardiomyocytes, respectively [97].
Oxidative stress, MAPK network, PKC and Ca2+ signaling activation in various viral infections including SARS-CoV and SARS-CoV-2 has already been described [98][99][100][101][102][103][104][105]. Ang II-mediated AT1R stimulation activates these signaling pathways, as well [106][107]. All these signaling pathways directly promote ADAM17-dependent ACE2 shedding which by itself decreases angiotensin(1–7)/Ang II concentration ratio and promotes Ang II/AT1R pathway which also induces ADAM17 sheddase function (see previous section). Additionally, COVID-19 has been shown to promote ADAM17 expression both at the protein and transcriptional level [108].
In this context, neither should it be forgotten that the downregulation of mACE2 in COVID19 can be due to internalization of the virus-ACE2 complex [109][110][111][112] nor should the cleaving effects of TMPRSS2 along with ADAM17 be ignored. These two proteases compete in shedding, yet at different sites, of ACE2: ADAM17 and TMPRSS2 cleave Arginine and Lysine amino acids in residues 652–659 and 697–716, respectively [113]. However, TMPRSS2-mediated dissociation of mACE2 does not result in the release of sACE2 [114].