International Union of Basic and Clinical Pharmacology. XCIX. Angiotensin Receptors: Interpreters of Pathophysiological Angiotensinergic Stimuli

1. Pharmacophore Structure-Activity Relationship.

Structurally, the biphenyl tetrazole seen in losartan, EXP3174, ZD7155, and candesartan or an acidic moiety as seen in telmisartan, 5-oxo-1,2,4-oxadiazole, as in azilsartan change lipophilicity of ARBs (Vyas and Ghate, 2010). Eprosartan has a different structure, biphenyl-tetrazole replaced with benzoic acid. Consequently, ARBs are designed to exploit hydrophobic interaction of the phenyl rings with transmembrane core of AT₁ receptor and bind basic residues in AT₁ receptor forming ionic interactions with the acidic moieties. Losartan has an imidazole with Cl at the other end of the molecule, and its metabolite, EXP3174, has a COOH substituent. Sartans present a greater variety of structures at this end that probably account for some of their functional differences. Olmesartan and losartan are closely related. Irbesartan has a cyclopentyl ring incorporated in place of the Cl. A benzimidazole is substituted in candesartan and azilsartan; two benzimidazoles are attached in telmisartan to produce an unusual structure for this ARB. The unique structure of valsartan lacks a nitrogen containing heterocycle. The imidazole ring in eprosartan has a large alkyl chain substituent. The difference in structure is expected to cause the ARBs to bind in slightly different ways.

In the crystal structure of AT₁ receptor bound to ZD7155, the ligand-binding pocket shows exactly how ARBs might interact with AT₁ receptor (Fig. 3). The AT₁ receptor residues, mainly from helices I, II, III, VII, and ECL2, directly interact with the ligand. Side chains of Arg^167(ECL2) and Tyr^35(TM1) form ionic and polar interactions with ZD7155. The acidic tetrazole moiety closely interacts with the basic Arg^167(ECL2), and this residue may be primarily responsible for ARB-binding affinity and selectivity for AT₁ receptor. Tyr^35(TM1) and the naphthyridin-2-one moiety of ZD7155 form hydrogen bonds. Trp^84(TM2) of AT₁ receptor forms π-π interaction with the naphthyridin-2-one moiety of ZD7155. Additionally, residues Ile^31(TM1), Val^108(TM3), Leu^112(TM3), and Tyr^292(TM7) in AT₁ receptor ligand-binding pocket interact hydrophobically with ZD7155. Most of the other contacts for ZD7155 binding to AT₁ receptor, however, are mediated by residues, including Tyr^87(TM2), Thr^88(TM2), Ser^105(TM3), Ser^109(TM3), Ala^163(TM4), Phe^182(ECL2), Pro^285(TM7), and Ile^288(TM7) (Zhang et al., 2015).

Shape of the binding cavity is determined by the secondary structures and disulfide crosslinking patterns of the extracellular loop region, proline and nonproline kinks in 7TM helical bundle, and other local variations, resulting in deviations in the extracellular tips of TM helices. Specific structural variations create a remarkable variety of sizes, shapes, and electrostatic properties of the ligand-binding pockets in different GPCRs. The ZD7155 binding site in AT₁ receptor partially overlaps with known ligand binding sites in the chemokine and opioid receptors in which the cavity is larger, more open, and located closer to the extracellular surface, a general feature of GPCRs that recognize diffusible small-molecule ligands. Some of the residues that comprise the ligand-binding pockets are conserved among these structurally similar peptide GPCRs, including Tyr^(TM1) and Trp^(TM2), and the majority of the residues forming the ligand binding of these receptors are close to the extracellular boundaries of the helices. Arg^167(ECL2) is one of the three critical residues that holds the antagonist close to the extracellular side and it is a unique residue of AT₁ receptor compared with other structurally similar peptide GPCRs.

Docking simulations of the clinically used antihypertensive ARBs show that they bind in similar orientations and interact with the three critical residues, Arg^167(ECL2)_, Trp^84(TM2), and Tyr^35(TM1) in the AT₁ receptor ligand-binding pocket (Fig. 3). Details of the interactions with residues Phe^77(TMII), Tyr^87(TMII), Ser^105(TMIII), Val^108(TMIII), Ser^109(TMIII), Leu^112(TMIII), Ala^163(TMIV), Phe^182(ECL2), Ile^288(TMVII), and Tyr^292(TMVII), which shape the ligand-binding pocket of the AT₁ receptor, are different for different ARBs. For example, one of the common features among these ARBs is a short alkyl tail with 2–4 carbons extending into a narrow hydrophobic pocket formed by Tyr^35(TM1)_, Phe^77(TM2), Val^108(TM3), Ile^288(TM7), and Tyr^292(TM7). Losartan is a surmountable antagonist with lower binding affinity to AT₁ receptor compared with the later developed ARBs (Takezako et al., 2004; Miura et al., 2011). Docking results suggest that Arg^167(ECL2) forms a salt bridge with only the tetrazole moiety of losartan but lacks other polar interaction and its interaction with Tyr^35(TM1); distances and angles for hydrogen bonding are suboptimal. The lower binding affinity and surmountable property of losartan for AT₁ receptor is likely due to this. In contrast, candesartan docking results indicate that it forms optimal interaction with the pocket. Lys^199(TM5) may form an additional salt bridge with the tetrazole moiety and further stabilize candesartan binding. Crystal structure and docking analyses suggest that Lys^199(TM5) provides conformational heterogeneity in the AT₁ receptor; the amino group of this residue may reach the acidic moieties of ARBs by forming salt bridges (as in the cases of candesartan and telmisartan) or through water-mediated interactions with other ARBs (Zhang et al., 2015). The structure of nonpeptide antagonists mimics the contact points of AngII side chains, and their binding sites overlap with the AngII binding pocket in AT₁ receptor as also supported by mutagenesis and crosslinking experiments (Yamano et al., 1992; Ji et al., 1994; Noda et al., 1995b; Vanderheyden et al., 2000a; Takezako et al., 2004).

Current literature is dominated by studies on losartan, candesartan, valsartan, irbesartan, olmesartan, and azilsartan that share a common biphenyl-tetrazole scaffold. Two ARBs in clinical practice differ, however. Telmisartan (also known as BIBR 277) structure is unusual, with two benzimidazole moeities attached (Fig. 4A). Eprosartan (also known as SK&F 108566), is the other ARB with the most differentiated structure. Telmisartan and eprosartan treatment benefits are not fully characterized, but potential benefits due to differences in their structural features are possible. These ARBs interact Arg167^ECL2, Tyr35^TM1, and Trp84^TM2 similar to other biphenyl-tetrazole ARBs, and mutants of these residues dramatically reduced their binding. However, the three-dimensional structure modeling studies predicted that telmisartan and eprosartan bind, interacting with significantly different residues (Fig. 4B). Eprosartan extends interaction into the hydrophobic subpocket consisting of Ile288^TM7 and Tyr292^TM7. Alanine substitution of both Ile288^TM7 and Tyr292^TM7 specifically decreased eprosartan-binding affinities. Two consecutive benzimidazole moieties in telmisartan were predicted to make additional π-π contacts with Tyr92^ECL1 (Fig. 4B). Mutation of Tyr92^ECL1 to alanine, although mostly neutral for other ARBs tested, significantly lowered affinity for telmisartan. Furthermore, Ile288Ala mutation also has a discriminating effect on telmisartan binding. Structural analysis reveals a novel paradigm, molecular recognition through extended interaction with receptor subpockets. This may suggest possible difference in outcome from blockade of AT₁ receptor function using different ARBs (Zhang et al., 2015).

2. Inverse Agonism.

Classic competitive antagonism model does not easily explain pharmacological behavior of most ARBs, therefore, more recent models to describe drug properties have been adopted. It is now generally accepted that AT₁ receptor has small but functionally significant constitutive activity (Unal and Karnik, 2014), and perhaps most ARBs do reduce constitutive activity of receptor in the absence of agonist, a phenomenon called inverse agonism. Inverse agonism of most ARBs is detected using constitutively active mutants of AT₁ receptor created by site-directed mutagenesis (Unal and Karnik, 2014). Inverse agonism has been shown for EXP3174, olmesartan, telmisartan, valsartan, and azilsartan (Noda et al., 1996; Miura et al., 2003a, 2006, 2013; Feng et al., 2005; Bhuiyan et al., 2009; Ojima et al., 2011). Whether losartan possesses inverse agonist properties remains controversial, because it was noted in one report (Bhuiyan et al., 2009) but not in further studies (Miura et al., 2003b; Feng et al., 2005). Clinical relevance of inverse agonism of ARBs is still debated. Kiya et al. (2010) studied nephroprotective properties of olmesartan and its analog lacking inverse agonism in Dahl salt-sensitive rats. Olmesartan lowered urinary protein excretion by ∼25%, but its close structural analog that lacked inverse agonism did not show nephroprotective effect. Differential effect of inverse agonist ARBs on proteinuria remains to be systematically studied.

3. Insurmountable and Reversible Antagonism.

Most clinically used ARBs exhibit an atypical competitive and reversible interaction with AT₁ receptor. The AngII concentration-response curves in the presence of ARBs shift toward higher concentrations, but with reduced maximal response. In experimental settings, a very high concentration of AngII cannot overcome the ARB inhibition fully. Most clinically used ARBs harbor this behavior (van Liefde and Vauquelin, 2009), but insurmountable antagonism of ARBs remains a rather descriptive term. Site-directed mutagenesis studies to identify receptor mechanisms have not provided a satisfactory answer (Vanderheyden et al., 2000b; Verheijen et al., 2003; Van Liefde and Vauquelin, 2009). Insurmountable antagonism of candesartan (Noda et al., 1993; Ojima et al., 1997), olmesartan, or EXP3174 was shown to increase with duration of preincubation with the receptor (Mizuno et al., 1995). Antagonism lasted for hours upon washout for many ARBs, including candesartan, EXP3174, olmesartan, telmisartan, and azilsartan. In insurmountable antagonism, the slow dissociation rates from the receptor may suggest conformational effects of ARB binding with clinically relevant functional consequences. The dissociation half-lives of the hAT₁ receptor has been measured, and the order of dissociation was found to be telmisartan, olmesartan, candesartan, valsartan, and losartan, respectively, with t_1/2 of 213, 166, 133, 70, and 67 minutes (Vanderheyden et al., 2000b; Verheijen et al., 2000; Kakuta et al., 2005)

4. Biased Agonism.

The ability of a given receptor to selectively activate a signaling path compared with another is defined as “biased agonism” or “ligand-directed” signaling. Biased ligands, also known as functionally selective agonists, can involve a subset of a receptor's normal signaling repertoire by stabilizing different receptor conformational states than are stabilized by endogenous “unbiased” ligands (Kenakin, 2007; Violin and Lefkowitz, 2007; DeWire and Violin, 2011; Godin and Ferguson, 2012). Functional selectivity exhibited by several AngII analogs has forced redefinition of receptor agonism paradigm to allow discrete receptor activation states from ligands with different efficacy for individual responses (Kenakin, 2005; Galandrin et al., 2007; Urban et al., 2007; Hansen et al., 2008; Lyngso et al., 2009). AngII peptide modification led to the description of agonists, partial agonists, and antagonists (Miura et al., 1999; Miura and Karnik, 1999; Oliveira et al., 2007). Hormone AngII is a full agonist with pluridimensional efficacy for all signals activated by AT₁ receptor, whereas AngII analog such as [Sar¹,Ile⁴,Ile⁸]AngII (SII-AngII) is a biased agonist that does not activate G protein signaling but allows receptor phosphorylation and subsequent β-arrestin–mediated signaling (Kenakin, 2003, 2005; Urban et al., 2007). This discovery has opened the possibility for novel designer ligands that may activate only some of AT₁ receptor signals. This development holds great promise for the treatment of diseases such as HF and hypertension (Holloway et al., 2002; Wei et al., 2003; Daniels et al., 2005; Aplin et al., 2007a,b; Hansen et al., 2008; Shukla et al., 2008; Sauliere et al., 2012).

Insight into the pluridimensional efficacy of AngII-analogs was initially described by Thomas et al. (2000) and Holloway et al. (2002). These authors studied the IP3 signaling, MAPK signaling, phosphorylation, and internalization of AT₁ receptor in response to AngII and several AngII analogs. Internalization of AT₁ receptor-EGFP was unaffected by substitution of Tyr-4 and Phe-8 in AngII that abolish IP3 signaling (Holloway et al., 2002). On the other hand, substitution of Asp-1 with alanine and substitutions of Phe⁸ with alanine, isoleucine, diphenylalanine, and β-cyclohexylalanine significantly inhibited phosphorylation. The activation of MAPK was inhibited by Phe-8 substitutions, which did not equally inhibit IP3 production or receptor phosphorylation, indicating overlapping but distinct AngII-AT₁ receptor interactions. Yee et al. (2006) described three distinct activated states of AT₁ receptor based on the response of several AT₁ receptor mutants to AngII and SII-AngII. Three active states proposed include AngII-mediated G-protein signaling, AngII-mediated G-protein–independent ERK1/2 signaling, and II-AngII–mediated G-protein–independent ERK1/2 signaling (Yee et al., 2006), thus linking receptor ligand combinations to generation of observed functional selectivity. SII-AngII is extensively studied for engagement of GRKs, antiapoptotic signals, chemotaxis, cell growth, and proliferation (Hunton et al., 2005; Yee et al., 2006; Aplin et al., 2007a,b; DeWire et al., 2008). An analog such as TRV120027 is a 30-fold more potent β-arrestin–biased AT₁ receptor ligand that has been studied in vitro and in vivo in rats and dogs and is now in Phase II clinical studies for the treatment of acute HF. The preclinical data are reported to indicate that TRV120027 has the potential to support heart, vasculature, and kidney function in acute HF (Violin et al., 2010; Boerrigter et al., 2011). TRV120027 may block the effects of elevated AngII on the vasculature and kidney, while protecting or enhancing prosurvival and contractility signals in the heart.

As seen in the case of many GPCR antagonists, ARBs do not cause internalization of the AT₁ receptor and block agonist-induced internalization. Unbiased antagonism by losartan, telmisartan, and valsartan are reported in internalization and desensitization (Violin et al., 2010). However, candesartan, losartan, telmisartan, and valsartan allow substantial internalization of constitutively active AT₁ receptor mutants (Bhuiyan et al., 2010). This phenomenon may reflect hidden biased agonism by these ARBs, a phenomenon that needs to be rigorously evaluated. Instances of a compound acting as an antagonist for one response and an agonist for a different response mediated by the same receptor are known (Patel et al., 2012). These findings contrast with studies in which candesartan binding did not induce internalization of a GFP-AT₁ receptor in Chinese hamster ovary (CHO) cells (Le et al., 2005). However, these aspects are yet to be described for AT₁ receptor and its peptide as well as nonpeptide ligands.

1. G-protein–Mediated Signaling.

Multiple heterotrimeric G-proteins interact with the AT₁ receptor, including G_q/11, G_i, G₁₂, and G₁₃, leading to activation of downstream effectors including phospholipase C (PLC), phospholipase A, and phospholipase D (Shirai et al., 1995; Ushio-Fukai et al., 1999a; Higuchi et al., 2007). The AT₁ receptor-G_q/11-phospholipase Cβ (PLCβ) coupling results in inositol triphosphates (IP3) and diacylglycerol (DAG) (Yusuf et al., 2000) signals. IP3 causes release of Ca²⁺ from the intracellular store. Intracellular Ca²⁺ cycling is the primary trigger for excitation-contraction of both cardiac and vascular myocytes. In addition, cytosolic Ca²⁺ does trigger intracellular signaling through calcineurin/nuclear factor of activated in T-cells or the calcium/calmodulin-dependent protein kinase II (CamK II) cascades, which are relevant to hypertrophy (Heineke and Molkentin, 2006). Histone deacetylase-5 has emerged as an important substrate of CamK II. A scaffold protein, GIT1, mediates AngII-induced VSMC gene transcription via CamK II-dependent phosphorylation and activation of HDAC5 (Pang et al., 2008). DAG activates PKC (Vallega et al., 1988), which also contributes to the vasoconstrictive and growth promoting effects of AngII. Phospholipase D–mediated phosphatidylcholine to choline and phosphatidic acid production is considered a second wave of signal by AngII activated AT₁ receptor. PA is rapidly converted to DAG, leading to sustained muscle contraction (Mehta and Griendling, 2007). AT₁ receptor causes activation of phospholipase A2 by phosphorylation and production of arachidonic acid and its metabolites. This signal is important in maintaining a balance between vasoconstriction and vasodilation in various vascular beds (Sarkis et al., 2004; Campbell et al., 1996) and NAD(P)H oxidation in VSMC (Griendling et al., 2000).

The G_βγ subunits released upon AT₁ receptor activation have been shown to activate tyrosine kinases such as SHC, pp60^c-src, and JAK2 (Gutkind, 1998; Gschwind et al., 2001; Kranenburg and Moolenaar, 2001; Luchtefeld et al., 2001), leading to downstream phosphatidylinositol 3′-kinase γ (PI3K_γ) activation (Lopez-Ilasaca et al., 1997). AngII-induced activation of AT₁ receptor specifically augmented G_β2 levels in the nucleus, where G_β2 interacts with specific nucleosome core histones and specific chromatin bound transcription factors such as myocyte enhancer factor 2 and thereby regulates functional gene networks (Bhatnagar et al., 2013).

2. G-Protein–Independent β-Arrestin-Mediated Signaling.

Multiple lines of observations led to the discovery that AT₁ receptor can directly recruit β-arrestin and mediate ERK1/2 signaling without G-protein activation (Holloway et al., 2002; Gaborik et al., 2003; Wei et al., 2003; Ahn et al., 2004a; Lee et al., 2008). The classic concept is that β-arrestin terminates G-protein signaling by blocking the receptor and stimulating receptor desensitization and endocytosis. However, β-arrestin–bound receptor can engage in G-protein–independent signaling, leading to distinct cellular responses (DeWire et al., 2007; Violin et al., 2013). This mode of signaling is recognized in a number of in vitro and in vivo settings; AT₁ receptor recruited β-arrestin1/2 engage a wide range of cellular responses.

The AT₁ receptor coupled β-arrestin mediates delayed ERK1/2 signaling confined to the cytoplasm for prolonged time (t_1/2 > 25 min). In contrast, both nuclear and cytoplasmic localization of activated ERK is G-protein–mediated (Gaborik et al., 2003; Tohgo et al., 2003; Ahn et al., 2004a). The AT₁ receptor-β-arrestin complex inhibits ERK-dependent transcription by binding to phospho-ERK and its retention in the cytosol (Tohgo et al., 2002). The cytoplasmic pool of ERK1/2 failed to phosphorylate the transcription factor Elk-1 and increase transcription of the immediate-early gene c-Fos (Tohgo et al., 2002; Aplin et al., 2007b). The β-arrestin2–dependent ERK activation may be responsible for distinct physiologic endpoints (Wei et al., 2004), such as increased survival of cardiac myocytes without entailing pathogenic myocyte hypertrophy. The beneficial effects of the β-arrestin–dependent pathway on enhancing cardiomyocyte survival was confirmed in a physiologic study in which a transgenic mouse with cardiac-specific overexpression of AT₁ receptor with second intracellular loop mutations that prevent G-protein coupling showed marked ventricular dilation and eccentric hypertrophy accompanied by diminished cardiomyocyte apoptosis in comparison with mice overexpressing a wild-type AT₁ receptor (Zhai et al., 2005).

Mechanistic details of activation of β-arrestin signaling by AT₁ receptor are emerging. AT₁ receptor can bind both β-arrestin1 and 2 (Oakley et al., 2000). Ahn et al. (2004b) demonstrated that physiologic levels of β arrestin1 might antagonize β-arrestin2–mediated ERK activation. Zimmerman et al. (2012) showed that AngII analogs selectively promoted β-arrestin–dependent effects in VMSCs. A mass spectrometry-based proteomics approach discovered that β-arrestin interacts with several proteins selectively after stimulation of the AT₁ receptor, indicating its potential for signaling (Xiao et al., 2007). In another study, activation by the β-arrestin biased ligand SII-AngII showed that unique phosphoproteins are activated (Christensen et al., 2010; Xiao et al., 2010). There are 34 differentially phosphorylated proteins, of which 16 were unique to SII-AngII and 8 were unique to AngII stimulation. Bioinformatics analysis of phosphorylated sites on these proteins identified downstream protein kinases activated by the AT₁ receptor-β-arrestin signaling (Kendall et al., 2011; Bogebo et al., 2014).

3. Reactive Oxygen Species Signaling.

AngII is a potent mediator of oxidative stress and oxidant signaling (Ushio-Fukai et al., 1999b; Taniyama and Griendling, 2003; Yan et al., 2003a; Touyz, 2004). AngII activates membrane NAD(P)H oxidase mainly via AT₁ receptor and PKC to produce reactive oxygen species (ROS) like superoxide and hydrogen peroxide (H₂O₂) (Rajagopalan et al., 1996; Ushio-Fukai et al., 1996; Zafari et al., 1998; Griendling et al., 2000; Seshiah et al., 2002; Touyz et al., 2005). ROS are involved in many pleiotropic effects of AngII such as activation of signaling molecules (e.g., c-Src, EGFR, p38MAPK, Akt) and transcription factors [e.g., nuclear factor κB (NFκB), NF-κB, AP-1, Nrf2], which are involved in atherosclerosis pathology (Sen and Packer, 1996; Chen et al., 2006; Papaiahgari et al., 2006; Wu et al., 2005). In endothelial cells, a low amount of superoxide production by eNOS maintains a supply of NO, contributing to vasodilation and vascular health. In disease state, eNOS is uncoupled, leading to excessive superoxide that reacts with NO to form peroxynitrite, a toxic radical, which directly contributes to disease (Schena et al., 1999; Zhao et al., 2005a; Taguchi et al., 2011).

4. Nonreceptor Type Tyrosine Kinase Signaling.

Nonreceptor tyrosine kinases associate with AT₁ receptor and target several intracellular proteins for phosphorylation. C-Src is a key player in AngII-mediated cellular effects. C-Src tyrosine kinase activity is also activated by ROS and it is involved in sustained calcium release (Sadoshima, 1998). Src is activated by AngII in an AT₁ receptor mutant lacking G-protein coupling ability, indicating that the carboxyl terminus of the AT₁ receptor is required for activation (Seta et al., 2002) because C-terminal truncation (310–359) abolished Src activation. VSMC growth is mediated by hyperactivation of c-Src-ERK1/2–dependent pathways by AT₁ receptor, leading to c-fos and AP-1 DNA-binding activity (Touyz et al., 2001b). AngII-stimulation of human monocyte migration is c-Src–dependent (Ishida et al., 1999) formation of focal adhesion complexes in the actin cytoskeleton (Polte et al., 1994). AT₁ receptor induces tyrosine phosphorylation of FAK, which then form a complex with Pyk2, p130Cas, paxillin, and talin, all of which interact to enable activation of cytoskeletal proteins facilitating adhesion of cells to extracellular matrix, and regulation of cell shape and movement (Leduc and Meloche, 1995; Sabe et al., 1997; Cary et al., 1998; Sayeski et al., 1998; Eguchi et al., 1999; Kintscher et al., 2001). Pyk2 is another kinase activated in response to AT₁ receptor and has been implicated in the regulation of ion channels, cellular adhesion, cell growth, and mitogenic and hypertrophic reactions (Sabri et al., 1998; Tang et al., 2000; Taniyama et al., 2003).

AT₁ receptor activates the JAK/STAT pathway via scaffolding by Src homology phosphatase-2 (SHP-2), a tyrosine phosphatase. The conserved YIPP motif in the AT₁ receptor interacts with SHP-2 and helps docking JAK2 and stimulates JAK2 phosphorylation at Tyr¹⁰⁰⁷/Tyr¹⁰⁰⁸ (Marrero et al., 1995a; Doan et al., 2001; Frank et al., 2002; Godeny et al., 2007). Upon activation of JAK2 by AT₁ receptor, STAT proteins are activated in order to mediate gene transcription of early growth response genes, such as c-fos and c-myc (Berk and Corson, 1997; Ishida et al., 1999; Luttrell et al., 1999; Madamanchi et al., 2001). Another tyrosine phosphatase, SHP-1, causes JAK2 dephosphorylation and termination of the AngII-induced JAK/STAT signaling (Marrero et al., 1998). A calcium/PYK2 and PKC pathway also exists for JAK2 activation in VSMCs (Frank et al., 2002). An AT₁ receptor mutant with tyrosine 292, 302, 312, 319, and 339 mutated to phenylalanine uncoupled from G-proteins was found to still activate tyrosine kinases and phospho-STAT1 signaling, indicating G-protein independence of these signals. JAK2 activation induces the expression of SOCS-3, which, in turn, blocks further activation of the pathway and consequently leads to desensitization of this signaling path (Torsoni et al., 2004).

5. Activation of Small G-protein Signaling.

AngII also activates family of small G-proteins, such as Ras, Rho, and Rac through the AT₁ receptor (Ohtsu et al., 2006c), which regulate the MAPK cascades in cardiovascular remodeling induced by AngII. Ras is activated in both cardiac myocytes and VSMCs (Eguchi et al., 1996; Sadoshima and Izumo, 1996) by SHP-2 phosphorylation, which leads to Shc/Grb2 complex and recruitment of the guanine nucleotide exchange factor, son of sevenless. Subsequently son of sevenless activates the Ras/Raf/MEK/ERK1/2 pathway and c-fos transcriptional activity (Sugden and Clerk, 1997; Ohtsu et al., 2006c). PKC can also associate with Ras and activate ERK1/2 (Liao et al., 1996; Liao et al., 1997). Accumulating evidence suggests that the Rho/ROCK (Rho-kinase) pathway is crucial for AngII-induced remodeling of vasculature (Aoki et al., 1998). Rho pathways are involved in cell migration (Seko et al., 2003). Rac is implicated in activation of p21-activated kinase 1 by AngII in VSMCs, which subsequently mediates JNK activation and hypertrophy (Seko et al., 2003; Woolfolk et al., 2005; Ohtsu et al., 2006a). Rho pathway is involved in the Ca²⁺ sensitization of smooth muscle contraction, pre-myofibril formation, and expression of atrial natriuretic factor in cardiac myocytes. Rac is an important component of the reduced NADPH oxidase complex to produce ROS by AngII in VSMCs (Gregg et al., 2003). ERK1/2, JNK, and p38MAPK activated downstream of AT₁ receptor-NADPH pathway are implicated in VSMC differentiation, proliferation, and migration (Sugden and Clerk, 1997; Taniyama et al., 2004), as well as in fibrosis and target-organ damage (Ishida et al., 1998; Ishida et al., 1999). The phosphatase MAPK phosphatase-1 (MKP-1) serves as a negative feedback control, inactivating ERK1/2 (Bokemeyer et al., 1998).

6. Transactivation of Receptor Tyrosine Kinase Signaling.

AT₁ receptor-induced transactivation of platelet derived growth factor receptor (PDGFR) modulates cell growth and migration (Heeneman et al., 2000; Suzuki and Eguchi, 2006) in VSMCs and mesangial cells (Linseman et al., 1995; Mondorf et al., 2000). This response is blocked by losartan and other ARBs. Upon acute AngII infusion, activation of PDGFR occurs in the vasculature of mice and rats. ACE-inhibitor infusion reduced aortic PDGFR phosphorylation and ERK activity (Kim et al., 2000), implicating PDGFR as downstream modulator of hypertensive vascular remodeling in vivo (Linseman et al., 1995; Heeneman et al., 2000; Mondorf et al., 2000; Gao et al., 2006).

AngII infusion also leads to activation of epidermal growth factor receptor (Zhang et al., 2009) in the vasculature (Kim et al., 2000). A major mechanism by which AngII influences growth-signaling pathways is through transactivation of EGFR by AT₁ receptor. AngII-induced renal deterioration involves EGFR transactivation mediated by ADAM17 (Lautrette et al., 2005) and second messengers such as Ca²⁺ and ROS (Diaz-Rodriguez et al., 2002; Mori et al., 2003; Seals and Courtneidge, 2003; Fischer et al., 2004; Tanaka et al., 2004; Mifune et al., 2005; Ohtsu et al., 2006a; Zhang et al., 2006). AT₁ receptor activation enhances the release of heparin-binding epidermal growth factor, which is dependent on A Disintegrin and Metalloproteinase (ADAM) family metalloproteinases (ADAM17) and Src (Andreev et al., 2001; Eguchi et al., 2001; Uchiyama-Tanaka et al., 2001; Schafer et al., 2004; Shah et al., 2004; Blobel, 2005; Mifune et al., 2005; Ohtsu et al., 2006a,b). Furthermore, dominant-negative ADAM17 mutant distinctly inhibited VSMC hypertrophy, which was stimulated by AngII (Ohtsu et al., 2006b). Heparin-binding epidermal growth factor activates EGFRs, allowing autophosphorylation on tyrosine (Prenzel et al., 1999). AngII-induced EGFR transactivation requires ROS and upstream kinases, such as c-Src, c-Abl, or Pyk2 (Dikic et al., 1996; Bokemeyer et al., 2000; Seshiah et al., 2002; Gratton et al., 2004) and leads to activation of the Ras/Raf/ERK pathway. AngII activation of Akt/PKB, p70S6K, and p38MAPK and induction of c-Fos leading to growth and migration of VSMCs, survival, and remodeling are mediated by EGFR (Che and Carmines, 2002; Seshiah et al., 2002; Suzuki et al., 2005).

A β-arrestin–dependent mechanism for transactivation of EGFR by AT₁ receptor has been reported (Kim et al., 2009). Calcium-independent AngII pathways also cause EGFR transactivation (Murasawa et al., 1998; Wang et al., 2000a) and the biased AngII analog [Sar¹,Ile⁴,Ile⁸]AngII activated G_q-independent EGFR signaling in human coronary artery smooth muscle cells (Miura et al., 2004). It was also reported that phosphorylation of tyrosine 319 and the YIPP motif of the AT₁ receptor is required for transactivation of EGFR (Seta and Sadoshima, 2003; Zhai et al., 2006). Cardiac-specific overexpression of an AT₁ receptor with a mutation in the YIPP motif (Tg-Y319F) did not cause cardiac hypertrophy in transgenic mice, and expression of fetal-type genes was significantly lower in these mice. Infusion of AngII failed to induce hypertrophy in Tg-Y319F mice, also significantly less apoptosis and fibrosis was reported (Zhai et al., 2006; Smith et al., 2011). EGFR activation is reported to be necessary for AngII-mediated hypertension and left ventricular hypertrophy (Ahmad et al., 2009).

7. Signaling through AT₁ Receptor Interacting Scaffold Proteins.

AT₁ receptor signaling occurs through the recruitment of scaffolding regulatory proteins. The carboxyl-terminal cytoplasmic region of the AT₁ receptor recruits different proteins to regulate different aspects of AT₁ receptor physiology (Mogi et al., 2009; Horiuchi et al., 2012). AT₁ receptor-associated protein (ATRAP1) is a transmembrane protein expressed in various tissues including the kidney, aorta, heart, lung, testis, and at a lower level in the lung, liver, spleen, and brain. It interacts specifically with the C-terminal tail and enhances AngII-induced internalization of AT₁ receptor (Daviet et al., 1999; Cui et al., 2000; Guo et al., 2003; Oshita et al., 2006; Azuma et al., 2007). ATRAP1 is a negative regulator of classic G-protein signaling by AT₁ receptor (Lopez-Ilasaca et al., 2003; Tamura et al., 2007) and VSMC growth (Cui et al., 2000) and cardiomyocyte hypertrophy (Tanaka et al., 2005). ATRAP1 prevents VSMC senescence (Wislez et al., 1998; Guo et al., 2005). ATRAP1 transgenic mice exhibited decrease in cardiac hypertrophy, neointima formation, inflammatory response, and NADPH oxidase activity in the injured artery (Oshita et al., 2006; Wakui and Tamura, 2012). In contrast, ATRAP1-deficient (ATRAP1−/−) mice showed increased mean systolic blood pressure and plasma volume, which was associated with increased surface expression of AT₁ receptors in the renal cortex and increased proximal tubular function (Oppermann et al., 2010). Kidney-specific ATRAP1 transgenic mice exhibit hypertension and renal hypertrophy and failure (Oppermann et al., 2010), suggesting that renal ATRAP1 plays an important role in regulating intrarenal RAS. Mechanistic aspects of AT₁ receptor interaction with ATRAP1 and transfer of signal are not clear at this time (Cook et al., 2008).

8. Mechanical Stretch.

AT₁ receptors were identified as mechanosensors in the myocardium by Issei Komuro's team by demonstrating agonist-independent activation of AT₁ receptor when stretch-stress is applied (Hunyady and Turu, 2004; Yasuda et al., 2008a; Mederos y Schnitzler et al., 2011). Mechanical stretch activated ERKs in the cardiomyocytes prepared from both neonatal and adult angiotensinogen-deficient mice (Zou et al., 2004), which could be inhibited by inverse agonist ARBs, such as candesartan. Mechanical stretch induces Janus kinase 2 and translocation of G-proteins into the cytosol (Zou et al., 2004; Yasuda et al., 2008a). The conformational changes in mechanically activated AT₁ receptor have been mapped (Shyu et al., 2001; Karnik et al., 2003; Yasuda et al., 2008b).

Sadoshima et al. (1993) initially reported that mechanical stretch causes secretion of AngII from cytoplasmic storage granules in cultured cardiac myocytes and that stretch-induced hypertrophic responses are completely dependent on the secreted AngII. However, several studies later showed that AngII partly mediates mechanical stress-induced hypertrophic responses (Yamazaki et al., 1995; Kijima et al., 1996). Therefore, mechanical stretch seems to directly activate unique intracellular signaling molecules. Yasuda et al. (2008b) showed that cell stretch leads to activation of the AT₁ receptor, which could be suppressed by candesartan, an inverse agonist. Thus, mechanical stress can directly change the conformation of the AT₁ receptor to increase the receptor’s basal activity. This was followed by Zou et al. (2004), demonstrating that the AT₁ receptor can be activated by mechanical stress inducing cardiac hypertrophy in vivo in an AGT-null background.

Stretch-induced activation of the AT₁ receptor protects cells against induced apoptosis involving PKB/Akt signaling (Kippenberger et al., 2005). Mechanical stretch potentiates AngII-induced VSMCs proliferation in spontaneously hypertensive rat through an AT₁ receptor/EGFR/ERK-dependent pathway. These findings may provide new insights into growth-promoting mechanisms in vasculature in a hypertensive state (Liu et al., 2010). Mechanical stretch triggered an AT₁ receptor-dependent conformational change in β-arrestin similar to that induced by a β-arrestin–biased ligand (Rakesh et al., 2010). These findings were unique to the AT₁ receptor (and not seen with the β1 adrenergic receptors) and suggest that AT₁ receptor is able to sense membrane stretch and transmit the activated receptor signal to β-arrestin.

9. Signaling through Heterodimerization.

The current view is that homo- and heterodimer formation of GPCRs could be important for some of the receptor functions (Lyngso et al., 2009). For example, the AT₁ receptor dimerizes with the bradykinin B2 receptor (Fig. 6), which enhances AngII signaling (AbdAlla et al., 2000, 2001b, 2005), contributing to AngII hypersensitivity in women with preeclampsia (AbdAlla et al., 2000, 2001b). The arrestin-biased ligand [Sar¹,Ile⁴,Ile⁸]AngII negatively regulates AT₁ receptor-B2R heterodimers by promoting sequestration of AT₁ receptor-B2R heterodimers (Wilson et al., 2013). Heterodimerization of AT₁ receptor with the MAS receptor and AT₂ receptor decreases AT₁ receptor-specific signaling (AbdAlla et al., 2001a; Kostenis et al., 2005; Canals et al., 2006; Santos et al., 2007). The AT₁ receptor can also form complexes with the β2 adrenergic receptors, and it is possible to effectively block dual receptor signaling using only a single receptor antagonist (Barki-Harrington et al., 2003). The AT₁ receptor also coimmunoprecipitates with the epidermal growth factor receptor (EGFR); dopamine D1, D3, and D5; and the endothelin B receptors (Zeng et al., 2003a,b, 2005a,b, 2006; Olivares-Reyes et al., 2005).

Evidence that AT₁ receptors can form dimers is supported by BRET analysis, suggesting homo- or oligomeric complexes in living cells that are unaffected by both agonists and antagonists (Hansen et al., 2004). Coexpression of signaling-deficient mutants results in functional receptors. AT₁ receptor wild-type G-protein coupling was diminished when coexpressed with defective mutant receptors, indicating “cross-inhibitory” association (Karip et al., 2007). Covalently crosslinked homodimer formation for AT₁ receptor is reported in isolated monocytes (AbdAlla et al., 2004) induced by factor XIIIA transglutaminase involving Gln³¹⁵ in the carboxyl-terminal tail of the AT₁ receptor. Hypertensive patients have increased homodimer levels (AbdAlla et al., 2004; Ogawa and Glass, 2004). Aldosterone produces a nongenomic endothelium-independent vasoconstrictor effect by enhancing intracellular transglutaminase activity and presumably inducing AT₁ receptor dimer formation in mesenteric arterioles perhaps due to transglutaminase-induced AT₁ receptor dimer formation (Yamada et al., 2008).

10. AngiotensinII Type 1 Receptor Signaling by Phosphorylation, Desensitization, and Internalization.

G-protein signaling by activated AT₁ receptor is accompanied by rapid phosphorylation and internalization of AT₁ receptor (Thomas, 1999; Hunyady et al., 2000; Guo et al., 2001; Thomas and Qian, 2003). Defects in desensitization are implicated in vascular diseases; for example, hypertensive rats overexpress GRK 5, altering AngII responsiveness (Ishizaka et al., 1997). AngII-induced desensitization of AT₁ receptor is dependent on carboxyl-terminal residues 329–347 (Conchon et al., 1998). This region is Ser and Thr rich, and phosphorylation of these residues plays a key role in the desensitization of AT₁ receptor responses. AngII-induced phosphorylation of the AT₁ receptor is mediated by both PKC and G-protein–coupled receptor kinases (GRK), mainly GRK2 and GRK5 (Oppermann et al., 1996a,b; Smith et al., 1998a; Qian et al., 1999). Mutation of the key serine and threonine residues in 332–338 region significantly inhibits AT₁ receptor internalization (Hunyady et al., 1994; Thomas et al., 1995, 1998; Smith et al., 1998b; Qian et al., 2001). Mutating hydrophobic residues in helix VIII in C terminus also inhibits AT₁ receptor internalization (Thomas et al., 1995). A diacidic motif of Asp²³⁶-Asp²³⁷ in the ICL3 of the AT₁ receptor is required for optimal AngII-induced phosphorylation of AT₁ receptor by GRKs and internalization (Olivares-Reyes et al., 2001). Complete deletion of the cytoplasmic tail inhibits internalization of AT₁ receptor (Chaki et al., 1994; Hunyady et al., 1994; Balmforth et al., 1995), and an STL motif (Ser³³⁵-Thr³³⁶-Leu³³⁷) in this region plays a critical role but also requires residues Leu³¹⁶ and Tyr³¹⁹ (Hunyady et al., 1994; Thomas et al., 1995).

Internalization of the AT₁ receptor follows both β-arrestin–dependent and –independent mechanisms (Lefkowitz, 1998; Somsel Rodman and Wandinger-Ness, 2000; Kim et al., 2005a). Alanine substitution for Thr³³², Ser³³⁵, Thr³³⁶, and Ser³³⁸ preclude agonist-induced β-arrestin recruitment by AT₁ receptor and attenuated internalization (Luttrell et al., 2001; Qian et al., 2001; Kule et al., 2004). Dramatically reduced AT₁ receptor internalization was observed in mouse embryonic fibroblasts lacking both β-arrestin 1 and β-arrestin 2 (Kohout et al., 2001). β-Arrestins target AT₁ receptor to clathrin-coated pits by interacting with clathrin and the clathrin adapter 2 (AP2), which directly interact with the AT₁ receptor carboxyl terminus (Fessart et al., 2005). These interactions are regulated by c-Src and the ADP-ribosylation factor 6 (ARF6), a small GTP-binding protein (Fessart et al., 2005; Poupart et al., 2007; Zimmerman et al., 2009).

After endocytosis, the receptor induces specific β-arrestin–mediated cell signaling pathways, distinct from G-protein signaling (Kim et al., 2005a). Shah et al. (2002) showed that β-arrestin–mediated ERK activation is regulated by transactivation of the EGFR and activation of GRK 5 and GRK6 (Kim et al., 2005a) but not GRK 2 and 3. The selective receptor phosphorylation on different sites by the various GRK isoforms may have important implications. For example, GRKs 5 and 6 seem to mediate effects associated with physiologic consequences, which are different from GRK 2–mediated effects (Kim et al., 2005a). Inhibition of GRK 5 or 6 attenuates β-arrestin–mediated ERK activation, whereas it is not affected by knockdown of GRK 2 or 3 (Kim et al., 2005a). An internalization-deficient mutant of the AT₁ receptor with truncated carboxyl terminus can also produce these responses, implying that internalization per se is not necessary for β-arrestin–mediated signaling (Turner et al., 2001).

AT₁ receptors are internalized within 10 minutes, and 25% of internalized receptors are recycled back to plasma membrane and the remainders are degraded in lysosomes (Gunther et al., 1980; Griendling et al., 1987). The Rab family of proteins, specifically Rab 1, is associated with transport of AT₁ receptor from endoplasmic reticulum to Golgi to cell surface (Wu et al., 2003). AT₁ receptor traffic to early endosomes is dependent on Rab 5 interaction with the carboxyl terminus of the receptor (Daviet et al., 1999; Somsel Rodman and Wandinger-Ness, 2000; Seachrist et al., 2002). Rab 7-positive late endosomes promote AngII dissociation. Rapid receptor recycling back to the plasma membrane takes place in Rab 11-positive vesicles by a rapid PI3K-dependent pathway (Garcia-Caballero et al., 2001; Hunyady et al., 2002; Seachrist et al., 2002; Dale et al., 2004).

In addition to clathrin-dependent pathway, AT₁ receptor can be internalized via specialized microdomains called caveolae, associated with caveolin (Ishizaka et al., 1998). In VSMCs, AngII regulates the expression, biosynthesis, and phosphorylation of caveolins and promotes the translocation of AT₁ receptor to caveolin-enriched membrane fractions (Ishizaka et al., 1998).

1. Cardiovascular Remodeling and Hypertrophy.

In vitro and in vivo experiments have shown growth promoting actions of AngII, causing cardiac and vascular hypertrophy, cell differentiation, and apoptosis (Pfeffer and Braunwald, 1990; Lombardi et al., 1999; Lips et al., 2003; Dorn and Force, 2005). In general, ARBs effectively prevent cardiac, vascular, and renal hypertrophy (Kim et al., 1995, 1998). Involvement of ERK1/2, PI3K, and CDK2 inhibition pathways, leading to G1-phase arrest, causes myocyte hypertrophy (Braun-Dullaeus et al., 1999). Increase in protein synthesis involves activation of translation elongation factor-2 in cardiac myocytes via dephosphorylation by PP2A by a process that involves both PI3K and MAPK (Everett et al., 2001). Increased protein synthesis through AT₁ receptors in human cardiac fibroblasts did not induce hypertrophy of cardiac fibroblast (Hou et al., 2000). In the vasculature, DNA synthesis is enhanced upon AngII infusion through the activation of cyclin D1 and cdk4 and reduction in the expression of cell cycle kinase inhibitors p21 and p27 (Diep et al., 2001; Guillemot et al., 2000, 2001). Thus, in the myocardium, regulation of growth effects by AT₁ receptor in myocytes and fibroblasts differs. Cardiac hypertrophy includes cardiac myocyte enlargement and proliferation of cardiac fibroblasts. Now it is generally believed that both hypertrophic response of myocytes and proliferative response of fibroblast may depend on other modifying factors such as production of ROS and secretion of various types of factors.

For instance, in pressure overload due to hypertension and myocardial infarction, cardiac remodeling process includes cardiomyocyte hypertrophy, extracellular matrix synthesis, fibrosis, and loss of compliance, leading to fatal outcomes. TGF-β1 expression is increased in myocytes and fibroblasts of heart, which transdifferentiate into a myofibroblast phenotype, resulting in myocardial remodeling (Campbell and Katwa, 1997). The AT₁ receptor directly increases TGF-β1 expression (Kupfahl et al., 2000; Schultz et al., 2002), translocation of Smad proteins Smad 2 and 4 into the nucleus, resulting in expression of fibrotic marker proteins, collagen, fibronectin, and connective tissue growth factor (CTGF) (Hao et al., 2000; Rodriguez-Vita et al., 2005; Lim and Zhu, 2006; Zhang et al., 2009). CTGF is a profibrotic factor that stimulates both AngII- and TGF-β-1–mediated fibrosis and apoptosis (Abreu et al., 2002; Perbal, 2004; Cabello-Verrugio et al., 2011). CTGF is involved in myocardial remodeling mediated via AT₁ receptors during transition to HF (Iwanciw et al., 2003; Ahmed et al., 2004). AngII-induced CTGF production is also shown in the aorta of AngII-infused rats (Ruperez et al., 2003). AngII-stimulated collagen synthesis in aortic adventitial fibroblasts, which is actively involved in vascular remodeling, is mediated by CTGF (Che et al., 2008). Gene expression analysis detected high CTGF mRNA expression in coronary artery biopsies from ischemic injury and coronary artery disease. Proteins involved in extracellular matrix remodeling, such as thrombospondin 4, collagen type 1 and 2, and fibronectin, and the inflammatory cytokines, such as IL-8, IL-6, vascular cell adhesion molecule-1, and monocyte chemoattractant protein-1 (MCP-1), are believed to couple cardiac remodeling with chronic angiotensin receptor stimulation (Gabrielsen et al., 2007). IL-6 secretion by cardiac myocytes is regulated by AngII. The effects of IL-6 on cardiomyocyte hypertrophy and fibroblast proliferation is inhibited by the AT₁ receptor antagonist losartan, suggesting that IL-6 contribution to cardiomyocyte hypertrophy is mediated by the AT₁ receptor (Fredj et al., 2005). AngII-induced activation of the JAK/STAT pathway is involved in tissue remodeling after vascular injury and myocardial ischemia in rats (Seki et al., 2000; Omura et al., 2001). The AT₁ receptor activates STAT1 and GATA4 transcription factors in the development of myocyte hypertrophy (Wang et al., 2005). Chronic activation of the AT₁ receptor in myocytes induces transcription of the Stat3 gene by pSTAT3 and overproduction of STAT3 protein, leading to nuclear accumulation of STAT3 without tyrosine phosphorylation, which alters the transcriptional program of cardiac hypertrophy (Yue et al., 2010).

VSMC hypertrophy induced by AngII involves PKC delta activation through Src-dependent Tyr phosphorylation, leading to Akt activation and signifying a novel molecular mechanism for enhancement of cardiovascular diseases induced by AngII (Nakashima et al., 2008). Antiapoptotic effects of AngII in cardiomyocytes and VSMCs are regulated by a mechanism involving PI3-kinase/Akt activation, subsequent upregulation of survivin, and suppression of caspase-3 activity (Ohashi et al., 2004). Inhibitors of Akt and a dominant-negative mutant of Akt selectively block AngII-induced proliferation of CHO-AT_1a cells (Dugourd et al., 2003). AngII activated reactive oxygen species acting through Src/caveolin-EGFR signaling pathway induces epithelial-to-mesenchymal transition in renal epithelial cells. This may be a novel molecular mechanism involved in progressive renal injury caused by chronic exposure to AngII (Chen et al., 2012). Adenoviral-directed expression of the AT₁ receptor has defined the EGFR transactivation pathway for cardiac hypertrophy via PI3K/Akt signaling (Ebert et al., 1995; Thomas et al., 2002).

2. Vascular Inflammation and Atherosclerosis.

The role of AngII in atherosclerosis has been well established. Atherosclerotic risk factors such as hypercholesterolemia and hypertension also increase production of angiotensinogen (Daugherty et al., 2004). Inhibition of AT₁ receptor s by losartan prevents lipid peroxidation, decreasing atherosclerotic lesion formation n apolipoprotein E–deficient mice (Keidar et al., 1997). Conversely, AngII infusion increases aortic atherosclerosis and aneurysm formation, independent of blood pressure (Daugherty et al., 2000; Weiss et al., 2001; AbdAlla et al., 2001a). Male apoE/AT_1a receptor double knockout mice have reduced atherosclerosis (Wassmann et al., 2004a). This relationship between AT₁ receptor and atherosclerosis is conserved in hypercholesterolemic rabbits (Yang et al., 1998). Stimulation of inflammatory mediators including IL-1β, IL-6, and TNFα through NFκB activation (Sanz-Rosa et al., 2005) and RAS activation appears to be the pathogenic mechanism in the atherosclerotic process. AngII and/or IL-6 infusion induces oxidative stress and endothelial dysfunction in mice and these effects are completely abolished in AT₁ receptor knockout mice (Schieffer et al., 2000; Ruiz-Ortega et al., 2001b; Wolf et al., 2002; Skurk et al., 2004; Wassmann et al., 2004b). In VSMCs, the activation of NAD(P)H oxidase by the AT₁ receptor produces IL-6 (Marui et al., 1993; Chen et al., 1998; Kranzhofer et al., 1999) and AT₁ receptor blockers decreases TNFα, IL-6, and soluble adhesion molecules (Tsutamoto et al., 2000). Secretion of inflammatory factors, such as P-selectin and MCP-1 that are involved in the vascular inflammation and atherogenesis is AT₁ receptor-dependent, and ARBs attenuate P-selectin and MCP-1 expression with concurrent reduction in intimal proliferation in mice (Chen et al., 2001). In hepatocytes, CARMA3/Bcl10/MALT1-dependent NFκB activation mediates AngII-responsive inflammatory signaling promoting pathologic liver fibrosis (McAllister-Lucas et al., 2007). CARMA1 and 3 are expressed in VSMC and endothelial cells, suggesting the potential for pathway similar to hepatocytes playing a role in vascular inflammation and atherosclerosis. The AT₁ receptor stimulates myofibroblasts to proliferate through activation of the NFκB transcription factor via a signaling pathway composed of CARMA3, Bcl10, and MALT1 (McAllister-Lucas et al., 2007).

Under chronic liver injury, AngII promotes pathologic liver fibrosis by stimulating hepatocytes and hepatic stellate cells to synthesize extracellular matrix proteins and secrete secondary cytokines. In rats, experimental models of liver fibrosis induced by bile duct ligation or carbon tetrachloride or choline deficiency, administration of irbesartan, olmesartan, telmisartan, candesartan, or losartan inhibited expression of collagens and TGFβ in stellate cells and reduced established liver fibrosis (Kurikawa et al., 2003; Ueki et al., 2006; Hirose et al., 2007; Yoshiji et al., 2009; Moreno et al., 2010; Kato et al., 2012). In clinical practice, however, the usefulness of treating liver fibrosis with ARBs remains contradictory (Schneider et al., 1999; Gonzalez-Abraldes et al., 2001; Lee, 2014). A handful of randomized controlled trials suggest that ARB treatment is a potentially useful therapeutic approach in patients with nonalcoholic fatty liver disease (Paschos and Tziomalos, 2012). The nonalcoholic fatty liver disease patients benefited from telmisartan and olmesartan treatment (Enjoji et al., 2008; Colmenero et al., 2009). In two independent pilot studies, administration of candesartan or losartan was shown to have an antifibrotic effect on patients with chronic hepatitis C (Sookoian et al., 2005; Ueki et al., 2009).

3. Endothelial Dysfunction.

Endothelial dysfunction refers to impairment of endothelium-dependent vasodilation associated with progressive changes in cell adhesion and barrier function. Inactivation of NO by AT₁ receptor–induced ROS is at the center of endothelial dysfunction in hypertension, atherosclerosis, and cardiovascular diseases. In endothelial cells, eNOS produces NO and a low amount of superoxide to maintain vasodilation and a healthy vasculature. In disease state, redox-uncoupled eNOS is responsible for disproportionate production of superoxide relative to NO. The excessive superoxide reacts with NO to form peroxynitrite, a toxic radical, which directly contributes to disease state. Endothelial cells in a disease state express adhesion molecules, which increase adhesion of circulating inflammatory blood cells into the vessel wall, leading to further increase in ROS production (Rajagopalan et al., 1996; Schena et al., 1999; Zhao et al., 2005a). ROS increases local RAS activation in the vasculature, enhancing AngII production in ECs, VSMCs, and fibroblasts and creates a mutual reinforcement loop between local RAS and ROS. The AT₁ receptor induces LDL receptor expression in ECs (Gryglewski et al., 1986; Rubanyi and Vanhoutte, 1986; Li and Mehta, 2000) and enhances the modification of LDL and the expression of its lectin-like receptor (LOX-1), which are critical events in atherosclerotic lesion formation (Chen et al., 2000a; Figueroa and Vijayagopal, 2002). Overall endothelial dysfunction is characterized by blunted endothelium-dependent vasodilation associated with enhanced contraction in hypertensive diseases (Dasgupta and Zhang, 2011).

4. Oxidative Stress.

AT₁ receptor–regulated oxidative stress plays a major part in the initiation and progression of hyperlipidemia, diabetes mellitus, hypertension, ischemic heart disease, and chronic HF (Marui et al., 1993; Taniyama and Griendling, 2003). ROS activates nuclear factor κB (NFκB) and stimulates degradation of its cytoplasmic inhibitor, IκB, an essential step in the proinflammatory process (Pueyo et al., 2000). Blockade of NFκB ameliorates myocardial hypertrophy in response to infusion of AngII, and the effect of AngII was attenuated in mice with targeted disruption of the p50 subunit of NFκB (Kawano et al., 2005). NFκB gene expression results in increased levels of vascular cell adhesion molecule-1 and probably other genes involved in the early stages of atherosclerosis (Pueyo et al., 2000). AngII/AT₁ receptor/ NFκB pathway may be involved in brain ischemia by stimulating intercellular adhesion molecule-1 expression in brain microvascular ECs (Liu et al., 2006). Thus, AT₁ receptor-induced ROS production can change structure-function properties of the vasculature that is the central aspect of vascular pathology in hypertension and diabetes. Treatment with ARBs stimulates NO release in platelets and ECs indicating arterial antithrombotic effects of ARBs (Kalinowski et al., 2002).

5. Extracellular Matrix Deposition.

AngII is a risk factor implicated in cardiac remodeling, and ARB treatment is protective (Nagata et al., 2002). AT₁ receptor signaling is increasingly recognized for its profibrotic effects in other tissues as well. Mechanism of fibrosis may differ in different tissues. In the heart, the AT₁ receptor upregulates TGF-β1, laminin, and fibronectin expression and contributes to increased cardiac fibroblast attachment to collagens I and III and increased focal adhesion kinase activity. Synthesis of the extracellular matrix proteins upon AT₁ receptor activation (Kato et al., 1991; Mifune et al., 2000) involves transactivation of EGFR-MAPK–dependent pathways (Ju and Dixon, 1996; Touyz et al., 2001a). Abnormal accumulation of proteoglycans is known in atherosclerotic lesions (Evanko et al., 1998; Iozzo, 1998), and treatment with ARBs induces proteoglycan changes that favor healthier cell adhesion, migration, and differentiation (Iozzo, 1998; Moriguchi et al., 1999; Sasamura et al., 2001). In aortic VSMCs and ECs as well as in cardiac cells, the AT₁ receptor–EGFR transactivation pathway also regulates fibronectin synthesis, production of matrix metalloproteinases and breakdown of collagen IV, expression of plasminogen activator inhibitor-1 (Feener et al., 1995; Chen et al., 2000b; Kawano et al., 2000; Nakamura et al., 2000). Regulation of PAI-1 by the AT₁ receptor seems to be important in many contexts; for instance, in human adipocytes, impairment of the fibrinolysis has been implicated in obesity (Skurk et al., 2001) and in reduced trophoblast invasion (Xia et al., 2002; Abbasi et al., 2005). AT₁ receptor blockade effectively reduces AngII-stimulated PAI-1 secretion (Sironi et al., 2001). In skin wound healing, AT₁ receptor–stimulated keratinocyte and fibroblast migration mediated by EGFR transactivation (Yahata et al., 2006) causes an increase in TGF-β1 and integrin protein levels (Thibault et al., 2001). This wound-healing pathway is attenuated in AT₁ receptor-knockout mice. Thus, the AT₁ receptor regulates formation of ECM components and turnover of matrix. The mechanisms and pathways that integrate ECM formation and turnover in relation to AngII signaling are still being discovered.

6. Insulin Resistance.

Patients with an imbalance in RAS homeostasis exhibit decreased insulin sensitivity (Nickenig et al., 1997; Kurtz and Pravenec, 2004), and treatment with ARBs improves insulin resistance and diabetic complications (Henriksen et al., 2001; Igarashi et al., 2001; Kurtz and Pravenec, 2004). Studies in rats show that AT₁ receptor activation hinders insulin signaling upon infusion of AngII to cause insulin resistance (Patiag et al., 2000; Ogihara et al., 2002). Normally, insulin binding to insulin receptor enhances its tyrosine kinase activity and tyrosine phosphorylation of the insulin receptor substrates (IRS) and activation of phosphatidylinositol-3 kinase (PI3K). In rat VSMCs, AngII impairs coupling of the insulin receptor (IGF-1R) to PI3K and inhibits insulin-mediated IRS-1 tyrosine phosphorylation and association of IRS-1 with p85 (Folli et al., 1997). Alternatively, AngII increases serine phosphorylation of IRS-1 (Ser616 via ERK and Ser312 via JNK), thus interfering with insulin signaling (Andreozzi et al., 2004). Another mechanism described for interference with insulin signaling involves AngII-dependent tyrosine phosphorylation of PDK1 and ROS-sensitive Src activation (Taniyama et al., 2004; Taniyama et al., 2005). PKC is another kinase that may interfere with insulin signaling (Motley et al., 2003). Thus, AT₁ receptor signals may hamper insulin action at multiple levels. Hypertension and diabetes are frequently seen together, indicating that interaction between AngII and insulin signaling plays an important role in cardiovascular pathology. ACE inhibitors are effective in the treatment of neuropathy found in diabetes (Malik et al., 1998; Malik, 2000). Coppey et al. (2006) showed that L-158809 (2-ethyl-5,7-dimethyl-3-[[4-[2-(2H-tetrazol-5-yl)phenyl]phenyl]methyl]imidazo[5,4-b]pyridine), an AngII receptor blocker, attenuated diabetic neuropathy in streptozotocin-induced diabetic rats. Until now, there are no clinical studies on the effect of ARBs in diabetic neuropathy, although the results suggest that ARB application may bear promise for treating neuropathy associated with vascular dysfunction and diabetic condition.

7. Angiogenesis and Cancer.

AT₁ receptor–regulated cell proliferation and angiogenesis has biologic and therapeutic implications in cancer (Escobar et al., 2004; Deshayes and Nahmias, 2005; Uemura et al., 2005, 2006, 2011; Ino et al., 2011; Lau and Leung, 2011; Uemura and Kubota, 2012). Local AngII production is a proangiogenic stimulus in tumor microenvironment. Tumors implanted in wild-type mice developed intensive angiogenesis with vascular endothelial growth factor (VEGF) induction in tumor stroma. Systemic administration of an ARB reduced tumor-associated angiogenesis and VEGF expression in tumor stroma. In comparison, tumor-associated angiogenesis was reduced in AT_1aR null mice, which was characterized by reduced expression of VEGF in the stroma and also reduced infiltration by macrophages. AngII-induced angiogenic factors production involves AT₁ receptor/JAK2/STAT3/SOCS3 signaling pathway. These results suggest that host stromal VEGF induction by AT₁ receptor is a key regulator of tumor growth and blockade of VEGF production by ARBs may be a novel therapeutic strategy against cancers (Egami et al., 2003; Fujita et al., 2005). AngII has been shown to function as a key role in neovascularization of hepatocellular carcinoma (Tamarat et al., 2002) in human breast carcinoma cells (Greco et al., 2002) as well as in invasive ductal breast cancer (Jethon et al., 2012). Both lisinopril and losartan treatment resulted in elevation in VEGF expression and angiogenesis, confirming the relationship between AT₁ receptor, VEGF, and vessel growth (Tamarat et al., 2002).

The choriocarcinoma cell proliferation is enhanced by AngII through the AT₁ receptor and activation of protein kinase C– and mitogen-activated protein kinase (Ino et al., 2003). AngII plays a role in the growth and chemoresistance of AT₁ receptor-positive pancreatic cancer cells through its action as a potent mitogen and antiapoptotic molecule (Amaya et al., 2004). AngII may promote prostate tumorigenesis via upregulation of PAX2 expression (Bose et al., 2009; Zhao et al., 2010). Thus, in the context of cancers of multiple tissues, AngII/AT₁ receptor signaling may favor tumor growth and may also contribute to aggressive etiology of cancers. Consideration of ARBs to restrict tumor growth should be a front line antitumor growth approach.

8. Autoantibodies and Malignant Hypertension.

Autoantibodies that bind to and activate the AT₁ receptor exist in patients with hypertensive disorders and contribute to disease pathophysiology in preeclampsia, in kidney transplant recipients who develop refractory vascular rejection, and in patients with malignant hypertension (Roberts, 2000; Lodwick, 2001; Dechend et al., 2004; Herse et al., 2008; LaMarca et al., 2011; Herse and LaMarca, 2013; Xia and Kellems, 2013). More recently, AT₁ receptor–directed autoantibodies have been seen in patients with the autoimmune diseases, including systemic sclerosis, featuring autoimmunity, vasculopathy, and tissue fibrosis (Fu et al., 2000; Liao et al., 2002; Ansari et al., 2005; Dragun et al., 2005; Riemekasten et al., 2011). The AT₁ receptor autoantibodies found in preeclampsia, renal allograft rejection, and malignant hypertension are directed to an epitope, -AFHYESQ-, in the second extracellular loop of the AT₁ receptor (Wallukat et al., 1999). Treatment with AT₁ receptor blocker reverses the pathophysiological effects of AT₁ receptor autoantibodies in these diseases, suggesting the ability of autoantibodies to activate AT₁ receptors (Dechend et al., 2000; Dorffel et al., 2003). Autoantibody-induced production of reactive oxygen species (ROS) by the placenta and maternal tissues likely contributes to the oxidative stress associated with preeclampsia (Hubel, 1999; Many et al., 2000; Sikkema et al., 2001; Dechend et al., 2003; Thway et al., 2004). Injection of pregnant mice with IgG from preeclamptic patients leads to hypertension, proteinuria, and preeclampsia in mice (Zhou et al., 2008b). These features were prevented by an antibody-neutralizing, seven-amino-acid epitope peptide (Zhou et al., 2007, 2008a). The presence of AT₁ receptor autoantibodies was also confirmed in the experimental rat models of preeclampsia as extensively reviewed by Xia and Kellems (2013) recently.

1. G-Protein Involvement.

All of the classic motifs and signature residues of a GPCR are present in the AT₂ receptor; however it fails to demonstrate classic features of G-protein signaling. Robust activation of heterotrimeric G-proteins, second messenger signals (calcium, DAG, cAMP, and IP₃), desensitization by phosphorylation, and receptor regulation by internalization is not observed in AT₂ receptor transfected cells (Kambayashi et al., 1993; Mukoyama et al., 1993; Miura and Karnik, 2002) or the native cells R3T3, PC12W, ovarian granulosa cells (Dudley et al., 1991; Pucell et al., 1991; Bottari et al., 1992a; Leung et al., 1992a; Webb et al., 1992). This behavior is consistent with observations of affinity of AngII binding, which is not affected by GTP or its analog GTPγS in tissues exclusively expressing the AT₂ receptor, such as human myometrium, rat adrenal glands, and bovine cerebellar cortex (Bottari et al., 1992a). Lack of agonist-induced desensitization and recycling of the AT₂ receptor is also observed in adult mice (Unger, 1999).

Multiple reports suggesting potential AT₂ receptor coupling to G_i/G_o have directly linked downstream signals to activation of this class of G-proteins (Fig. 7). For instance, AT₂ receptor s sensitive to GTPγS and pertussis toxin have been described in the locus coeruleus of the rat brain and thalamic or geniculate nuclei implying a coupling to heterotrimeric G-proteins, G_i and G_o (Tsutsumi and Saavedra, 1991). The AT₂ receptors appear to bind to Gi_α2 or Gi_α3 in developing fetus in rat (Zhang and Pratt, 1996) and in adult rat kidney cells. The intracellular third loop (ICL3) of the AT₂ receptor was shown to be important for its coupling to G_i (Hayashida et al., 1996). Evidence of ICL3 involvement was also shown in AT₂ receptor induction of apoptotic responses in the PC12W neuronal lineage cells (Lehtonen et al., 1999) and in AT₂ receptor–mediated inhibition of IP₃ generation in Xenopus oocytes (Kumar et al., 2002). More recent studies implicated the involvement of G_i in AT₂ receptor–dependent increases in nitric oxide synthase expression (Li et al., 2007a) and the inhibition of proximal tubule Na⁺-ATPase by Ang(1-7) (Siragy and Carey, 1996; Siragy et al., 1996; Gohlke et al., 1998; Siragy, 2000). The production of prostacyclin in differentiated adipocytes was blocked by PD123177 and not by losartan, which suggests that this is an AT₂ receptor–mediated signal (Darimont et al., 1994). Whether this mechanism involved a G-protein, a G_i signaling pathway remains to be clarified. Direct G-protein activation assays in heterologous expression systems remain difficult with the AT₂ receptor. Whether ICL3 is important for coupling of the full-length receptor to G_i has not been validated by mutations in ICL3. Discrepant findings regarding G-protein coupling with the AT₂ receptor is an unresolved research challenge that constrains a full understanding of AT₂ receptor signaling mechanisms.

2. Protein Phosphatase Involvement.

Depending on the tissues, activation of the AT₂ receptor can stimulate protein phosphatases (MKP-1, PP2A, SHP-1) and protein dephosphorylation (Fig. 7). The activation of vanadate-sensitive tyrosine phosphatase as well as Ser/Thr phosphatases has emerged as a key mechanism accounting for the antigrowth and apoptotic effects of the AT₂ receptor (Bottari et al., 1992a; Horiuchi et al., 1997; Elbaz et al., 2000). The activation of the protein tyrosine phosphatase SHP-1 mediated by AT₂ receptor is pertussis toxin insensitive, thus a G_i-independent signal (Bedecs et al., 1997). SHP-1 coupling may actually involve an atypical G-protein scaffolding mechanism, G_βγ-independent constitutive association of the receptor with G_s and SHP-1 (Feng et al., 2002). In contrast to these studies, a G_i-coupled activation of PTP by AT₂ receptor in VSMCs was reported. Activation of G_i/G_o in these cells by a peptide corresponding to the third intracellular loop of the AT₂ receptor was inhibited by pertussis toxin, as well as by sodium orthovanadate a tyrosine phosphatase inhibitor. The serine/threonine phosphatase inhibitor okadaic acid did not block it (Hayashida et al., 1996). Similarly, activation of soluble SH-PTP1 to attenuate MAPK activation and DNA synthesis through thymidine incorporation observed in NIE-115 neuronal cells was suggested to be G_i/G_o dependent.

Activation of PTPases by the AT₂ receptor in cells may limit mitogen-activated protein kinase (MAPK) signaling and dephosphorylation of extracellular signal-regulated kinases 1 and 2 (ERK1/2) (Huang et al., 1995; Hayashida et al., 1996; Bedecs et al., 1997; Fischer et al., 1998; Akishita et al., 1999). The AT₂ receptor is a generalized negative regulator of intracellular kinase signaling, and thus it may antagonize growth stimulation by AngII, EGF, PDGF, or serum. AT₂ receptor knockout mice confirm this aspect, because elevated levels of ERK1/2 were reported at baseline and in response to serum (Akishita et al., 1999). ERK1/2 dephosphorylation may occur through three phosphatases: SHP-1, mitogen-activated protein kinase phosphatase 1 (MKP-1), and protein phosphatase 2A (PP2A) (Huang et al., 1995; Horiuchi et al., 1997; Yang et al., 1999). The AT₂ receptor–mediated suppression of MAPK activation in vessels is reported (Nakajima et al., 1995). The carotid artery expresses low levels of the AT₂ receptor (Viswanathan and Saavedra, 1992), but re-expressed AT₂ receptor at the edge of neointima during carotid injury or in a healing wound of the skin suppresses AngII-induced MAPK activity in a PD123319-sensitive manner to facilitate remodeling of these tissues (Nakajima et al., 1995). AT₂ receptor activation is linked to upregulation of SHP-1 in vascular-targeted AT₂ receptor transgenic mice (Matsubara et al., 2001). The AT₂ receptor in the neonatal hypothalamic neurons is reported to inactivate MAPK, and the presence of the AT₂ receptor antagonist PD123319 enhanced MAPK activity in these neurons. The ERK inactivated by the AT₂ receptor plays a physiologic role in vivo, depending upon the biologic context in which AT₂ receptor is engaged. For instance, in PC12W cells, which express only AT₂ receptor, AngII-regulated short-lived ERK phosphorylation plays a role in neuronal differentiation by AngII and nerve growth factor costimulation (Stroth et al., 2000). In the heart of AT₂ receptor transgenic mice, the antihypertrophic state of myocytes is associated with ERK dephosphorylation, which indicates that ERK inactivated by the AT₂ receptor plays a physiologic role in vivo (Masaki et al., 1998).

Conflicting reports also exist regarding phosphatase activation in AT₂ receptor–mediated proapoptotic effects. Dephosphorylation of Bcl-2 by MKP-1 was observed in AT₂ receptor–induced apoptosis in PC12W cells, which was ligand dependent (Horiuchi et al., 1997); apoptosis was shown to be a constitutive function of the AT₂ receptor that involves activation of p38 MAPK (Miura and Karnik, 2000). Overexpression of the AT₂ receptor in neonatal cardiomyocytes promoted growth (D'Amore et al., 2005). Overexpression and AngII stimulation of the AT₂ receptor in porcine cardiac fibroblasts inhibited protein tyrosine phosphatases (Warnecke et al., 2001). Observed contradictions further highlight the context-specific effects of the AT₂ receptor overexpression.

3. Scaffolding Protein Involvement.

Choosing an unbiased approach to understanding how the AT₂ receptor suppresses cellular growth in different contexts yielded intriguing leads (i.e., yeast 2-hybrid “fishing” with the AT₂ receptor C-terminal tail as bait), which are indicative of completely novel signaling modalities (Fig. 7); however, these are in-progress studies that might lead to novel physiologic endpoints.

Negative regulation of several receptor tyrosine kinases (RTKs) including fibroblast growth factor, epidermal growth factor (Zhang et al., 2009) and insulin receptors by AT₂ receptor is described as transinactivation by AT₂ receptor activation of PTPases as well as inhibition of autophosphorylation of the RTKs (Elbaz et al., 2000; De Paolis et al., 2002). Possible mechanisms suggested for the transinactivation of RTKs include a direct physical interaction between the two receptors. The ErbB3 EGF receptor as an interacting partner with C terminus of the AT₂ receptor was identified. These authors also showed that replacing ICL3 of the AT₂ receptor with that of AT₁ receptor abolishes the interaction with ErbB3 and the inhibitory effects on cell proliferation and the activation of apoptosis (Pulakat et al., 1998, 2002).

The AT₂ receptor C-terminal tail interaction with the transcription factor promyelocytic zinc finger protein (PLZF) is one of the most surprising discoveries. After AngII stimulation, PLZF translocates from the cytosol to the plasma membrane and then promotes internalization of the AT₂ receptor with PLZF, leading to accumulation of AT₂ receptor in the perinuclear membrane. PLZF enters the nucleus where it activates the p85α subunit gene of PI3K and enhances p70^S6 kinase activity, which is essential for protein synthesis (Senbonmatsu et al., 2003). The finding that AT₂ receptor internalizes when associated with PLZF is an interesting phenomenon, given that AT₂ receptor does not normally internalize (Mukoyama et al., 1995; Turu et al., 2006). PLZF is a highly expressed transcription factor in the heart; however, its expression is not affected in the AT₂ receptor–knockout mice. AT₂-knock–out mice do not show a hypertrophic response to pressure overload after aortic banding, this phenotype is associated with a failure to upregulate p70^S6 kinase after pressure overload. The AT₂ receptor–null hearts also fail to activate p85α transcription in response to AngII infusion. These data strongly suggest that the AT₂ receptor-PLZF-p85α-p70^S6 kinase–signaling axis might be important in the induction of cardiac hypertrophy (Senbonmatsu et al., 2000).

Two independent groups identified ATIP/ATBP50 (AT₂ receptor interacting protein) (Nouet et al., 2004; Wruck et al., 2005). At least four members are known to date for the ATIP-family (ATIP1-4). They all conserve an AT₂ receptor–interacting domain and can inhibit insulin, epidermal growth factor, and basic fibroblast growth factor-induced ERK1/2 activation and DNA synthesis in CHO cells in a similar manner to AT₂ receptor (Nouet et al., 2004). The ATIP-mediated inhibition of ERK1/2 requires AT₂ receptor expression but not treatment with AngII (Nouet et al., 2004). Li et al. (2007b) showed that siRNA-mediated knockdown of ATIP inhibits AT₂ receptor–mediated expression of methane methylsulfonate–sensitive 2, which plays an important role in the ubiquitin proteasome system and DNA repair. Methane methylsulfonate–sensitive 2 expression is increased in occluded/remodeling cerebral arteries, which is consistent with a protective role for the AT₂ receptor in brain injury. The siRNA-mediated knockdown of ATIP is also reported to reduce cell surface expression of the AT₂ receptor and suppress the antiproliferative effect (Wruck et al., 2005).

Although these studies yielded novel and interesting directions to pursue, the rationale for choosing the AT₂ receptor C-terminal tail as bait is not fully justified. There is no evidence to suggest that the AT₂ receptor C-terminal tail mediates the functions ultimately described for each scaffold protein. Rather surprisingly, involvement of G_s/G_i/G_o proteins and the ICL3 appear to be an important determinant of AT₂ receptor coupling to PLZF as well as for interactions with ErbB3. These findings again point to G-protein requirement at least in unconventional mode and raise doubts about the mechanisms of scaffold signaling by the AT₂ receptor .

4. Nitric Oxide/cGMP Involvement.

Nitric oxide (NO)–stimulated soluble guanylyl cyclase catalytic activity generates cGMP to generally exert protective effects in various tissues (Toda et al., 2007). Initial in vitro studies in neuronal cell lines suggested that AngII via the AT₂ receptor reduced cGMP levels in neuronal cells (Sumners and Myers, 1991; Sumners et al., 1991; Bottari et al., 1992b; Brechler et al., 1993). In other in vitro studies cGMP was not detected (Leung et al., 1992b; Webb et al., 1992; Mukoyama et al., 1993, 1995; Siragy and Carey, 1996). The studies by Liu et al. (1997) demonstrated that beneficial effects of AT₂ receptor involved kinin stimulation and cGMP production. Gohlke et al. (1998) also showed that AT₂ receptor stimulation increased cGMP levels in the rat aorta, which was further corroborated in AT₂ receptor–null mice studies (Siragy et al., 1999). The AT₂ receptor transgenic mice have elevated levels of cGMP in the aorta (Tsutsumi et al., 1999).

Whether the AT₂ receptor directly couples to NO/cGMP system especially in AT₂ receptor transgenic and knockout mice or through an indirect bradykinin B2-receptor dependent mechanism has been controversial. Tsutsumi et al. (1999) showed that the AT₂ receptor stimulates bradykinin production in VSMCs, and AngII-mediated cGMP response could be blocked with the bradykinin receptor antagonist. AT₂ receptor stimulates bradykinin production and promotes the NO/cGMP pathway in a paracrine manner, and this mechanism is important in AT₂ receptor–mediated vasodilatation of human coronary arteries (Batenburg et al., 2004). Different signaling circuitry may be involved in AT₂ receptor–dependent NO/cGMP production. In the thoracic abdominal aortic constriction mice model, AT₂ receptor activation induces phosphorylation of eNOS via a PKA-mediated signaling pathway (Yayama et al., 2006). Bradykinin acting on the bradykinin B₂ receptor can also induce PKA-dependent phosphorylation of eNOS. Evidences for a functional heterodimerization of the AT₂ receptor and B₂R receptors have been reported, which suggest that these receptors may physically associate and possibly increase NO production (Abadir et al., 2006).

5. Ion-Channel Protein Involvement.

A highly reproducible signal from the AT₂ receptor seems to be coupling to ion-channel proteins (Fig. 7), such as the hypothalamic neuronal delayed rectifier potassium channel (Kang et al., 1993, 1994), the T-type calcium channel (Buisson et al., 1992, 1995) in other cells, which can suppress or induce cellular growth and differentiation (Nakajima et al., 1995; Stoll et al., 1995; Laflamme et al., 1996; Meffert et al., 1996; Munzenmaier and Greene, 1996; Tsuzuki et al., 1996a,b; Gallinat et al., 1998; Stroth et al., 1998; Cote et al., 1999; Gendron et al., 1999) as well as support apoptosis (Yamada et al., 1996; Chamoux et al., 1999; Gallinat et al., 1999).

Cardiac ventricular myocytes express both the AT₁ receptor and AT₂ receptor. Expression of AT₁ receptor along with AT₂ receptor is increased in the hypertrophic ventricles of SHR and two-kidney one-clip hypertensive rats (Suzuki et al., 1993; Busche et al., 2000). AngII-stimulated arachidonic acid production in cardiac myocytes is fully blocked by the AT₂ receptor antagonist PD123317, suggesting that AT₂ receptor mediates the activation of phospholipase A2 (Lokuta et al., 1994). The AT₂ receptor–arachidonic acid signal activates the Na/HCO symporter system, which maintains pH in myocytes. The arachidonic acid–induced symporter activation in the heart is a unique function of the AT₂ receptor (Sandmann et al., 1998). In rats after myocardial infarction, the postinfarct upregulation of the Na1/H1-exchanger (NHE-1) and NBC ion transporter systems are differentially regulated by AT₁ and AT₂ receptors. ACE inhibitors prevent this regulation in heart tissue. However, the AT₁ receptor antagonists block NHE-1 upregulation, and the AT₂ receptor antagonists block the increase of the NBC (Horiuchi et al., 1999; de Gasparo et al., 2000; Nouet and Nahmias, 2000). This difference may be important in evaluating therapeutic intervention.

6. Involvement of Constitutive Activity.

Ligand-independent actions of the AT₂ receptor (Miura and Karnik, 2002) is recognized as an important mode of its in vivo function in several scenarios. Upregulated AT₂ receptor expression as seen in remodeling adult tissues and developing fetal tissue can initiate constitutive signal transduction without AngII stimulation, leading to apoptosis (Miura and Karnik, 2000). Stimulation of apoptosis in prostate cancer cells was mediated by increased expression of the AT₂ receptor (Li et al., 2009b). AT₂ receptor gene transfer in the same cells mediated increased expression of bradykinin and iNOS in VSMCs. Downregulation of the AT₁ receptor was also observed by the same treatment (Jin et al., 2002). In another gene transfer study, Porrello et al. (2009) found that the AT₂ receptor constitutively antagonized AT₁ receptor–mediated cardiomyocyte autophagy. Homooligomerization of the AT₂ receptor through disulfide exchange between Cys³⁵ in one AT₂ receptor and Cys²⁹⁰ in its dimerization partner was shown to be important for induction of apoptosis without AngII stimulation (Miura et al., 2005). Augmented release of bradykinin in mouse coronary artery endothelial cells was constitutive function (Zhu et al., 2010). Repression of ERK activity mediated by ATIP requires AT₂ receptor expression but not activation by AngII (Nouet et al., 2004). Feng et al. (2002) provided a potential mechanism for the constitutive function of the AT₂ receptor. They documented that constitutive association of G_α with SHP-1 and the AT₂ receptor is G_βγ independent and this association is essential in AT₂ receptor-mediated ITIM-independent activation of SHP-1.

At the cellular level, a clear mechanistic model of AT₂ receptor signal transduction and cell physiology has yet to emerge. Despite intensive investigation, the AT₂ receptor is one of the inadequately understood components of the renin-angiotensin system.

1. Developmental Regulation.

Human tissue expression of AT₂ receptor has been documented by immunocytochemistry, reverse transcriptase–polymerase chain reaction, and microarray analyses (Ichiki and Inagami, 1995; de Gasparo et al., 2000; Herradon et al., 2004; Kim et al., 2005b; Chakrabarty et al., 2008; Petracco et al., 2012). In most tissues, the AT₁ receptor and AT₂ receptor coexist. The AT₂ receptor mRNA expression analysis provided a better picture of the tissue- and species-specific distribution (Kakuchi et al., 1995; Shanmugam et al., 1995; Johren and Saavedra, 1996). The AT₂ receptor is abundantly and ubiquitously expressed in the developing rat fetus and neonatal tissues, which lead to early speculation of a developmental role for the AT₂ receptor (Suzuki et al., 1993; Bastien et al., 1996).

In mouse fetal mesenchymal cells, >95% of the AngII binding is due to the AT₂ receptor. Regulation of its expression in relation to fetal developmental stages may suggest a potential role for this receptor in developmental processes. AT₂ receptor expression levels increase at embryonic days E11–E13. Maximal AT₂ receptor expression levels are seen on E19 followed by a rapid decline in the newborn animals to undetectable expression. The AT₂ receptor–null genotype has only a marginal effect, if any, on the development of mesenchyme-rich tissue or organs in the mouse such as skin, tongue, kidney, or adrenal (Kakuchi et al., 1995). However, a high incidence of urological abnormalities (Graham et al., 1997) and delay in vasculogenesis was reported in AT₂ receptor null mice (Yamada et al., 1998). The AT₂ receptor is the exclusive AngII receptor in the fetal rat aorta, which decreases by ∼85% at 2 weeks of age and to 25% at 8 weeks of age (Viswanathan and Saavedra, 1992). AT₂ receptor blockade with PD123319 from E16 to E21 significantly decreases DNA synthesis in the developing aorta (Nakajima et al., 1995), and the AT₂ receptor–null mice have reduced levels of the VSMC differentiation markers calponin and caldesmon up to 4 weeks after birth (Yamada et al., 1999). In a similar fashion, stimulating the AT₂ receptor induces neurite outgrowth and regulates neurofilament expression in neural cells (Laflamme et al., 1996) as well as promotes apoptotic effects neuronal cell lines (Yamada et al., 1996; Horiuchi et al., 1997). Increase in neuronal cell number in certain brain structures associated with central neurologic abnormalities, learning, and memory deficit have been reported for AT₂ receptor knockout mice (Hein et al., 1995; Ichiki et al., 1995; von Bohlen und Halbach et al., 2001). Whether the increase in cell number in these mice is due to increased neuronal proliferation or a suppression of apoptosis is not clear. The proapoptotic effect of the AT₂ receptor is not found in the cardiomyocytes of AT₂ receptor transgenic mouse (Sugino et al., 2001).

2. Adult Tissue Regulation.

Receptor expression declines after birth, but the AT₂ receptor is expressed at low levels in the normal adult cardiovascular system, adrenal gland, kidney, brain, uterine myometrium, and skin. The capacity for regulated changes in expression of AT₂ receptor is retained in many adult tissues and often in remodeling tissues, a finding that suggests a potential role for this receptor in human cardiovascular disease as well.

The expression of both AT₁ receptor and AT₂ receptor in the adrenal gland, nonpregnant uterus, ovarian follicular granulosa cells, and cerebellar cortex is conserved in all mammals including humans (David, 1976; Whitebread et al., 1989; Criscione et al., 1990; Bottari et al., 1991; Pucell et al., 1991; Cox et al., 1993). There are some examples of tissues that contain more AT₂ receptor than AT₁ receptor. In the uterus, for example, expression of AT₂ and AT₁ receptors changes reversibly depending on pregnancy and parturition in human and sheep (Cox et al., 1993). The expression switching from high AT₂ receptor levels to a low-level AT₂ receptor and to completely AT₁ receptor is a fascinating developmental phenomenon that is not clearly understood.

The AT₂ receptor density increases in tissues under pathologic conditions in which inflammation and tissue remodeling occur. Such changes have been observed in tissues, such as skin wound, balloon-catheterized vasculature, infarcted myocardium, and ischemic brain during pathogenesis of hypertension, atherosclerosis, diabetes mellitus, nephropathy, and pulmonary fibrosis (Booz and Baker, 1996; Lemarie and Schiffrin, 2010). The observed increase in AT₂ receptor expression during pathology can be viewed as fetal gene reactivation, which is a prominent feature of various disease states, and this response may indicate that the AT₂ receptor has a definitive regulatory function in these instances (Paul et al., 2006). In rodents, AT₂ receptor expression is upregulated in HF (Ohkubo et al., 1997; Steckelings et al., 2005) and is up- and downregulated in a temporally dependent manner postinfarction (Nio et al., 1995; Lax et al., 2004). Expression of AT₂ receptors in humans is less clear. A decrease in AT₂ receptor expression during HF was reported (Matsumoto et al., 2000), whereas, an increase in expression was seen during atrial fibrillation (Goette et al., 2000). Myocardial AT₂ receptor expression is upregulated in patients with dilated cardiomyopathy. In the nonfailing human heart ∼40% of AngII receptors are AT₂ receptor (Tsutsumi et al., 1998), and the AT₂ receptor constitutes 50–70% of angiotensin binding sites in the adult myocardium in humans (Regitz-Zagrosek et al., 1995). In the adult rodent myocardium, ∼10% of adult cardiomyocytes express AT₂ receptor (Busche et al., 2000; Steckelings et al., 2005), but AT₁ receptor blockade increases AT₂ expression levels in rodents. Hence, under conditions of AT₁-inhibitor therapy, circulating AngII levels increase, creating the potential for increased AT₂ receptor function and facilitation of therapeutic tissue remodeling. This is the basis for targeting AT₂ receptors for therapy in patients treated with AT₁ receptor blockers to treat hypertension and HF. Adult brain AT₂ receptor may regulate central effects of angiotensin peptides in osmoregulation, cognitive functions, and behavioral functions (Hein et al., 1995; Hohle et al., 1995). Additionally, the role of the AT₂ receptor in regeneration and protection of neuronal tissue has been suggested (Steckelings et al., 2011a).

In the vasculature, a ratio of 80% AT₁ receptor to 20% AT₂ receptor exists in coronary endothelial cells derived from spontaneously hypertensive rats (SHR) (Stoll et al., 1995). Adult vasculature has low levels of AT₂ receptor, which may increase the sensitivity of vasculature to AngII when an AT₂ receptor-specific antagonist is present via the vascular AT₁ receptor (Verlander et al., 2011). The inhibitory effect of the AT₂ receptor on neointima formation after balloon catheterization of carotid artery has been extensively studied. The growth of the neointima was reduced by the AT₂ receptor antagonist PD123319 (Nakajima et al., 1995). Chronic AngII infusion increased the wall thickness-to-lumen ratio of mesenteric arteries associated with increase of both AT₁ receptor and AT₂ receptor (Cao et al., 1999). End-stage ischemic heart disease and dilated cardiomyopathy increased the density of the AT₂ receptor in endocardial, interstitial, perivascular, and infarcted regions of the ventricle of patients. That AT₂ receptor s may play a role in vessel formation is suggested by rich microvessels found in the border zone between noninfarcted and infarcted myocardium (Wharton et al., 1998).

In the kidney, 10 to 20% of total AngII receptors are AT₂ receptor, and no AT₂ receptors are detected in the rat renal cortex and glomeruli. Dietary sodium may modulate expression levels of AT₂ receptor in glomerular and interstitial tissue. The AT₂ receptor is present in large preglomerular vessels of the renal cortex and in the tubular interstitial cells in human kidney (Chansel et al., 1993; Goldfarb et al., 1994; Kaufman et al., 1999; Carey et al., 2001).

1. Regulation of Vascular Response.

Physiologic effects of AT₂ receptor function in the vasculature are complex and can be both vasoconstricting and vasodilatating depending on the context. AT₂ receptor activation causes vasodilatation in most isolated arteries and it exerts a vasodepressor effect in vivo, when a concomitant, low-dose AT1R blockade is also administered (Widdop et al., 2003). Actions of the AT₂ receptor antagonist PD123319 and agonist CGP42112 consistent with this expectation have been demonstrated in rodent arteries as well as in human coronary microarteries (Carey et al., 2001; Batenburg et al., 2004). The pressor response to AngII is greater in AT₂ receptor–null mice than normal controls. VSMC-targeted overexpression of AT₂ receptor in transgenic mice causes vasodilatation in vivo (Tsutsumi et al., 1999). That the AT₂ receptor also mediates vasoconstriction in mesenteric resistance arteries of SHR and senescent rats is a recent finding (You et al., 2005). Antihypertensive treatment of 4 weeks restored AT₂ receptor expression and vasodilator function in SHR resistance arteries. Pinaud et al. (2007) reported that the AT₂ receptor induces a vasodilator effect in young rats and vasoconstrictive effects in old rats. Resolving the apparently conflicting roles of the AT₂ receptor in different vascular beds should be a focus for future studies.

2. Regulation of Cardiac Growth Response.

The AT₁ receptor induces mitogenic effects in many tissues and cell types. Regulation of cell proliferation by the AT₂ receptor through an antiproliferative action opposed to the AT₁ receptor activation first discovered in coronary endothelial cells is also observed in microvascular endothelial cells, VSMCs, neuronal cells, pheochromacytoma cells, and fibroblasts (Stoll et al., 1995; Steckelings et al., 2005). The potential clinical ramifications of an AT₂ receptor-agonist therapy for the treatment of proliferative pathologies including cardiac fibrosis are vast. But the literature reports are confusing, for instance, AT₁ receptor–dependent stimulation of DNA and protein synthesis by AngII in neonatal cardiomyocytes and cardiac fibroblasts is observed only when the AT₂ receptor–selective antagonist PD123319 is present, suggesting that the cellular AT₁ receptor/AT₂ receptor ratio determines the outcome (Schorb et al., 1993; Crawford et al., 1994; van Kesteren et al., 1997). Increasing AT₂ receptor expression in neonatal cardiomyocyte cultures mediates myocyte hypertrophy independent of AngII (D'Amore et al., 2005), indicating that the AT₂ receptor is prohypertrophic. Studies of AT₂ receptor knockout and transgenic mice are ambiguous on the physiologic role of the AT₂ receptor in fibrosis. The myocardial perivascular fibrosis is increased in AT₂ receptor null mice (Wu et al., 2002), and an opposite outcome was observed in a different AT₂ receptor knockout mouse model (Kurisu et al., 2003). Overexpression of the AT₂ receptor in the heart significantly inhibited AngII-induced increases in perivascular fibrosis (Kurisu et al., 2003) in one model, whereas in another mouse model interstitial collagen increased (Yan et al., 2003b).

With regards to hypertrophy of myocardium, AT₂ receptor manipulation experiments in vivo yielded conflicting results. AT₂ receptor knockout prevents the induction of hypertrophy upon AngII infusion (Ichihara et al., 2001). However, transgenic mice overexpressing the AT₂ receptor in cardiomyocytes are normal and after AngII infusion they develop the same degree of hypertrophy as normal controls (Masaki et al., 1998; Moore et al., 2001; Yan et al., 2003b). Surprisingly the hypertrophic response to pressure overload is suppressed in AT₂ receptor–null mice, suggesting the AT₂ receptor is essential for hypertrophic process (Senbonmatsu et al., 2000). In a ventricular myocyte directed AT₂ receptor overexpression model, the AT₂ transgenic mice have an impaired Ca²⁺-dependent inotropic response to AngII, associated with reduced activity of the Na⁺/H⁺ exchange (Nakayama et al., 2005). Transgene copy number influenced cardiac hypertrophy in this model. Hypertrophy under basal conditions was observed in MLC2v-AT₂TG mice with 18 copies of transgene with highest level of AT₂ receptor expression, whereas mice with 9 transgene copies did not display any signs of cardiac hypertrophy or HF (Yan et al., 2003b). Subjecting the 9 transgene copy mice to pressure overload by aortic banding significantly reduced left ventricular myocyte diameter, systolic pressure, and collagen compared with aortic-banded nontransgenic controls (Yan et al., 2003b). Thus, the role played by AT₂ receptor in heart is essential for both AngII-induced and pressure-overload myocardial hypertrophy, again raising doubts about the “yin-yang” paradigm that this receptor is a physiologic antagonist of AT₁ receptor actions.

3. Regulation of Fibrosis Response in Other Tissue.

In vivo, PD123319 increases renal fibrosis, supporting the notion that the AT₂ receptor is anti-fibrotic in kidney as well. However, there are also conflicting reports that chronic PD123319 administration reduces collagen content in rats (Levy et al., 1996). Aggravated renal injury is observed in AT₂ receptor–knockout mice (Benndorf et al., 2009). The AT₂ receptor–deficient mice had exaggerated mortality accompanied with increased risk of albuminuria, renal fibrosis, glomerular injury, lymphocyte infiltration, and chemokine expression compared with renal ablation control mice. The blood pressure and RAS metabolites were similar in these groups.

Studies in disease models carrying the AT₂ receptor-null genotype show greater susceptibility for pathogenesis of cerebral infarction and atherosclerosis (Iwai et al., 2005). Angiotensin AT₂ receptor protects against cerebral ischemia induced neuronal injury (Li et al., 2005; Kaschina et al., 2008). In cerulean-induced pancreatitis, expression of AT₂ receptor is associated with pancreatic fibrosis, which is increased through a TGF-β–mediated mechanism (Ulmasov et al., 2009).

In the last 10 years, strides have been made toward understanding the role of the AT₂ receptor in various pathogenic models (Knowle et al., 2000; De Paolis et al., 2002; Miura and Karnik, 2002). The AT₂ receptor has been implicated in cardiovascular disease as a beneficial moiety due to its proposed vasodilatory role, ability to inhibit cell growth, induce apoptosis, and inhibit activation of MAPK functions that, overall, appear to counterbalance the actions of the AT₁ receptor (Berk, 2003; Miura et al., 2010). Overall, the AT₂ receptor may play a role in several AT₁ receptor–independent biologic processes (Mifune et al., 2000; Ruiz-Ortega et al., 2000, 2001a; Zhu et al., 2000; Weidekamm et al., 2002; Benndorf et al., 2003; Gingras et al., 2003; Caballero et al., 2004; Zhao et al., 2005b; Mertens et al., 2010). Furthermore, activation of the AT₂ receptor inhibits autophagy mediated by the AT₁ receptor in cardiomyocytes (Porrello et al., 2009). Notably, activation of the AT₂ receptor in cells of neuronal origin, induced neurite outgrowth and elongation, mediated cellular excitability, and migration and, in some cases caused, neuronal cell death (Guimond and Gallo-Payet, 2012). Taken together, it has become clear that the AT₂ receptor signaling mechanisms are diverse and require further studies to fully elucidate the role of this receptor in biologic processes.

1. Agonists of the AngIV Binding Site.

The characteristics of the AngIV molecule critical for high-affinity binding to the AngIV binding site in tissue include preservation of residues at positions 1, 2, or 3 of the AngIV, whereas positions 4, 5, and 6 can be replaced (Sardinia et al., 1993; Wright et al., 1995). A detailed study (see Supplemental Material) of position 1 of AngIV revealed that high-affinity AngIV binding requires a primary amine in position 1 and l-hydrophobic residues at position 1 favor binding. Thus, l-norleucine1-AngIV resulted in a very high-affinity analog (K_d in the picomolar range). The CH₂-NH isostere substituted at 1–2 peptide linkage produced a high-affinity analog that increases rotation of peptide bond (Sardinia et al., 1993). An l-aromatic amino acid at position 2 and l-hydrophobic amino acid at position 3 of the AngIV molecule are critical to bind to AngIV binding site. Deletion of the C-terminal phenylalanine from of AngIV does not alter binding affinity, but this derivative is inactive in manipulating physiologic blood flow (Wright et al., 1995). The minimal AngIV-like peptide ligand to bind AngIV binding site with high affinity requires a sequence, Val¹-Tyr²-Ile³-R1-R2. An agonist requires a phenylalanine in addition at the C-terminal position, a structural feature of AngIV, which would also permit activation of signaling through the ubiquitously present AT₁ receptor. In most studies of AngIV analogs, inadvertent activation of the AT₁ receptor may have to be carefully ruled out. The peptide LVV-hemorphin-7 binds and activates the AngIV binding sites in vivo. LVV-H7 specifically displaced [¹²⁵I]AngIV binding to adrenal and brain membranes.

2. Antagonists of the AngIV Binding Site.

Harding’s laboratory made early strides overcoming the difficulties of design and synthesis of AngIV binding site antagonists. They had success with Val substitution coupled with reducing peptide bonds in the Val-Tyr-Val segment generating divalinal-AngIV (Sardinia et al., 1993; Sardinia et al., 1994; Krebs et al., 1996). A second antagonist with norleucine in position 1, leucine in position 3, and reduced peptide bonds at the first and third bond positions (Norleual-AngIV) was subsequently created and characterized. This analog binds the AngIV binding site with high affinity and has been reported to not bind AT₁ receptor and AT₂ receptor. Divalinal-AngIV prevented c-Fos expression, blood flow induced by AngIV in brain, as well as the inhibition of proximal tubule sodium transport by AngIV (Handa et al., 1998).

3. Inhibitors of the Insulin Regulated Amino Peptidase Activity.

Concept of IRAP as the potential AngIV binding site stimulated development of newer and more selective analogs and resulted in the development of a new class of cognitive enhancers, which target the catalytic site of IRAP (Albiston et al., 2008) (see Supplemental Material). Efforts were taken to develop AngIV-derived IRAP inhibitors with improved selectivity versus AP-N and the AT₁ receptor, as well as resistance to proteolytic degradation. Lukaszuk et al., 2008) reported that the AngIV analog AL-11 (pK_i 7.25 for IRAP enzyme activity) in which the β-homo amino acid β²hVal is responsible for stability and the β³hPhe for selectivity. Further improvement was obtained by substitution of His⁴-Pro⁵ with a constrained Trp analog (Aia-Gly) to obtain AL-40 with a pK_i is 8.07 for IRAP (Lukaszuk et al., 2009). Substitution of the latter Gly⁵ by norvaline led to an analog (IVDE77) with further increased affinity (pK_i is 8.77 for IRAP) (Lukaszuk et al., 2011; Nikolaou et al., 2013). The latter compound was also ³H-labeled and characterized as a useful tool for the detection of IRAP/AngIV binding site under physiologic conditions, i.e., in the absence of any metal ion chelators (Nikolaou et al., 2013).

Synthesis of peptidomimetics targeting IRAP activity lead to successful replacement of the C-terminal His-Pro-Phe in AngIV replaced by small aromatic γ-turn mimetics (Compound 29, K_i = 26 nM). Compounds with low nanomolar inhibitory activity and enhanced selectivity for IRAP over the structurally related amino peptidase, AP-N, were obtained. Additional studies are needed to elucidate bioactive conformation for IRAP inhibition and presumably functional selectivity for AngIV binding site–mediated cognitive effects (Anderson et al., 2010, 2011; Axen et al., 2006). With a bioinformatics in silico screening approach, a large library of compounds against a structure homology of IRAP (Ye et al., 2008; Albiston et al., 2011) benzopyran compounds were identified, synthesized, and optimized; the highest affinity compound HFI-437 has a pK_i of 7.70 for IRAP. The latter nonpeptide inhibitors were found to enhance the performance of rats in different memory paradigms (Mountford et al., 2014).

Yang et al. (2008) used IRAP/AngIV binding site-selective ligands, LVV-hemorphin-7 and AT4-16, for intravenous, intrarenal, and intracerebroventricular infusion to discriminate AngIV effects through AT₁ and AngIV binding sites. They observed that AngIV had no effects on mean arterial pressure, renal blood flow, or urinary sodium excretion in the presence of IRAP inhibitors. In addition, they reported that the majority of the AngIV-sensitive aminopeptidase activity in kidney membranes is due to AP-N and not the IRAP/AngIV binding site. Some of the central and peripheral vasoactive effects of AngIV were mediated through the AngIV binding sites in other studies (Lew et al., 2003; Axen et al., 2007). Increase in blood pressure from chronic AngIV elevation in the brain was found to be AT₁ dependent in a transgenic mouse model, although specificity of action of AngIV in this model should be a question (Lochard et al., 2004). Thus, central and peripheral pressor and vasoconstrictor effects of AngIV are very likely mediated through AT₁ receptor in vasculature. The activation of the AT₁ receptor by AngIV is a well established pharmacological effect, but this effect is distinct from the memory-enhancing effects of AngIV (Yang et al., 2008). Stragier B. et al. (2006) independently demonstrated that AngIV and LVV-H7 did not significantly alter hippocampal blood flow as determined by laser Doppler flow measurement, excluding the vascular effects responsible for cognitive functions of either peptide (Jewett et al., 1991) (see Fig. 8 and Supplemental Material).

1. Effects on Insulin Regulated Amino Peptidase Signaling as a Transmembrane Receptor.

Prior to the identification of IRAP as AngIV binding site, the AngIV binding site was thought to be a multimeric, 225-kDa complex in bovine tissues (Bernier et al., 1995b; Zhang et al., 1999). It is now clear that AngIV binding sites are not coupled to G-proteins in tissues such as bovine endothelium, guinea pig brain, and rat vascular smooth muscle (Hall et al., 1992). The N-terminal domain of IRAP may participate in signal transduction with AngIV acting as an agonist in the activation of intracellular signaling, which may mediate the physiologic effects of AngIV. Vauquelin et al. (2002b) proposed that IRAP is a classic transmembrane receptor, with an intracellular tail that would interact with intracellular signaling molecules. This idea is based on the finding that AngIV interacts with a juxta-membrane domain of IRAP/AngIV binding site in bovine heart membranes. This putative binding site for AngIV is different from that of the active site of IRAP, which implies that AngIV binding may modulate transmembrane signal transduction by AngIV binding site and the IRAP activity by an allosteric mechanism (Caron et al., 2003). In the structurally related AP-N, binding of monoclonal antibodies activates IP3-linked increase of intracellular calcium and phosphorylation of MAPK (Navarrete Santos et al., 2000a,b,c). Expression of c-Fos is induced by intracerebroventricular infusion of AngIV in brain centers of rats, which are associated with cognition (Roberts et al., 1995). Several studies have shown that AngIV binding site ligands are capable of activating different intracellular signaling pathways. These include a variety of mechanism(s), including an increase of the intracellular calcium concentration, modulation of MAPK activation of NFκB signaling, increased cGMP production, and more downstream changes in DNA synthesis. It should also be noted that in no classic signaling could be demonstrated in CHO cells despite the presence of AngIV binding sites (Albiston et al., 2007). In pulmonary artery endothelial cells, AngIV binding to AngIV binding sites activates eNOS and stimulates cGMP accumulation and causes pulmonary arterial vasodilation (Patel et al., 1998). This effect was blocked by divalinal AngIV as well as by the phosphoinositide 3-kinase inhibitors (Patel et al., 1998). Induction of PAI-1 by AngIV has been implicated in the pathogenesis of renal interstitial fibrosis in several forms of chronic glomerulonephritides (Gesualdo et al., 1999). AngIV stimulates cell proliferation in rat anterior pituitary without requiring AT₁ receptor (Ptasinska-Wnuk et al., 2003) and also activates NFκB, leading to the transcription of proinflammatory genes, intercellular adhesion molecule-1, IL-6, MCP-1, PAI-1, and TNFα in VSMCs. These effects were blocked by treatment with AngIV binding site antagonist, suggesting that signal transduction activated by AngIV may underlie pathogenesis of cardiovascular diseases (Esteban et al., 2005).

IRAP inhibitors, such as divalinal-AngIV and LVV-hemorphin-7, which bind to the IRAP active site directly, produce similar effects. For instance, increasing the extracellular dopamine levels in the striatum of the rat by these inhibitors was comparable to AngIV. This finding suggests that active-site binding of the inhibitors mediates modulation of extracellular dopamine levels in striatum (Stragier et al., 2004). In a related study, the authors hypothesized that IRAP and/or AP-N possibly act as receptors for IRAP inhibitors and AngIV in mediating dopamine release in the striatum by receptor-mediated signaling (Stragier et al., 2007). High-affinity binding of AngIV to only the Zn-depleted IRAP apoenzyme may inhibit IRAP activity by stabilizing an inactive apoenzyme. AngIV binding to the catalytic domain of IRAP has been clearly established in vitro, and the evidence for AngIV binding to an allosteric site in the juxtamembrane region has been questioned. [¹²⁵I]AngIV binds in an a 1:1 ratio to in the presence and absence of the juxtamembrane region, suggesting that there is only one AngIV binding site but the site is not located in the juxtamembrane region. AngIV binding to the apoenzyme can be measured in vitro, reflecting possibly a different mechanism of binding and stabilization of AngIV when Zn is removed. Taken together, AngIV appears to bind better to the IRAP apoenzyme than the native IRAP. AngIV binding to native AP-N or to the AP-N apoenzyme is poor (Demaegdt et al., 2006), indicating that AP-N as an in vivo receptor for AngIV is poorly supported by data. Thus, redundancy of mechanism of AngIV receptors is possible.

2. Effects on Insulin Regulated Amino Peptidase-Dependent Neuropeptide Potentiation.

On the basis of the kinetics of inhibition, AngIV and its analogs have been proposed as in vivo inhibitors of neuropeptide catabolism by IRAP (Ye et al., 2007). AngIV and divalinal display competitive kinetics, indicating that AngIV binding site ligands mediate their peptidase inhibitor effects by binding to the catalytic site of IRAP (Lew et al., 2003). IRAP is known to play a role in the in vivo metabolism of vasopressin and Lys-bradykinin bound to the enzyme (Herbst et al., 1997; Wallis et al., 2007). Substrates identified in vitro for IRAP include vasopressin (AVP), oxytocin, somatostatin, Leu- and Met-enkephalin, and Lys-bradykinin (Herbst et al., 1997; Matsumoto et al., 2001; Lew et al., 2003). These neuropeptides enhance memory consolidation and retrieval (Vauquelin et al., 2002b; Wallis et al., 2007). The physiologic effects of AngIV binding site ligands therefore may result from the inhibition of cleavage of neuropeptides by IRAP altering memory processing (Lew et al., 2003). Accumulation of neuropeptides when AngIV binds to IRAP has been suggested as a mechanism for cognitive enhancement. In a recent study it was shown that the increase of extracellular dopamine concentration and the anticonvulsant effect of AngIV could be inhibited by a somatostatin receptor antagonist administered in the rat striatum (Li et al., 2009a). Accumulation of somatostatin to mediate these effects may be one possible mechanism for the observed phenomenon, although somatostatin cleavage by IRAP was not confirmed. At the moment there is no direct in vivo evidence that supports the IRAP/AngIV binding-mediated neuropeptide potentiation hypothesis.

3. AngIV Effects on IRAP Trafficking.

Although AngIV and its analogs are not cleaved by IRAP, they competitively inhibit the in vitro catalytic activity of IRAP. On this basis, it has been suggested that IRAP inhibitor binding may prolong the cell surface localization of IRAP/AngIV binding sites in neurons to potentiate cognitive and memory functions (Heimfeld et al., 1992; Demaegdt et al., 2004a,b). In tissues not directly involved in cognitive functions, insulin and some IRAP substrates peptide hormones have been shown to regulate kinetics of cell-surface trafficking of IRAP. For instance, in vascular endothelial cells in placenta, intracellular vesicle-bound IRAP responds to oxytocin by translocating the IRAP vesicles to the plasma membrane. By catabolizing oxytocin and vasopressin, translocated intracellular vesicle-bound IRAP regulates oxytocin and vasopressin levels in serum during pregnancy and thereby prevents early contractions and premature delivery. Cotranslocation of IRAP and the glucose transporter GLUT4 associated vesicles to the plasma membrane in adipocytes is stimulated by insulin (Demaegdt et al., 2011). The cytosolic N-terminal domain of IRAP binds p115, a protein that mediates transport vesicle fusion and is cotranslocated with IRAP and GLUT4 to the plasma membrane (Hosaka et al., 2005). The intracellular domain of IRAP plays a role in regulating GLUT4 vesicle trafficking through interaction with specific proteins including AS160 (Larance et al., 2005; Peck et al., 2006; Albiston et al., 2007). GLUT4 has a critical role in insulin-stimulated glucose uptake in these cells (Keller et al., 1995; Ross et al., 1996). The subcellular localization of IRAP in neurons demonstrated striking parallels with distribution of IRAP in insulin responsive adipocytes and muscle cells, where the enzyme plays a role in insulin-regulated glucose uptake. Therefore, the function of IRAP in neurons may be similar to that in insulin responsive cells (Fernando et al., 2007). Relevant to cognitive function of IRAP/AngIV binding site in brain, the majority of IRAP was localized in intracellular vesicles in mouse brain neurons (Fernando et al., 2005, 2007). The binding of IRAP inhibitors may prolong the cell surface localization of IRAP/AngIV binding site and GLUT4 in neurons. The resulting increase of glucose uptake may be responsible for IRAP ligands to facilitate memory in hippocampus-dependent memory deficits in models for amnesia (Fernando et al., 2007; Stragier et al., 2008). The IRAP inhibitors enhance cAMP-stimulated or depolarization-evoked glucose uptake mediated by GLUT4 (Fernando et al., 2008). Although both AngIV and LVV-H7 potentiate glucose uptake into cultured hippocampal neurons in culture, these effects were not confirmed in vivo. Extracellular hippocampal glucose levels significantly decreased while the animals explored the maze. When the animals were not in a maze, glucose levels in their hippocampus remained constant. However, the same analysis performed in the plus maze test showed that both AngIV and LVV-H7 did not significantly alter hippocampal glucose levels compared with control. Increased glucose availability for hippocampal neurons as the basis of enhanced memory seems unlikely.

1. Evaluation of AngIV-dependent Functions.

Considering the wide-ranging effects mediated by IRAP and its inhibitors, understanding the genetic basis of IRAP functions is critical. In normal and memory-compromised rodent models, enhancement of memory upon intracerebroventricular delivery of AngIV and its analog peptides is the defining pharmacological property, which is thought to be mediated solely by the IRAP/AngIV binding site (Albiston et al., 2004b; Lee et al., 2004). Biochemically, these ligands show high-affinity binding to the IRAP catalytic site. Therefore, the inhibition of IRAP activity or genetic loss of expression would be expected to improve memory.

Albiston et al. (2010) evaluated effects of IRAP gene knockout on AngIV binding and the behavioral phenotype in mice. In the IRAP-null mouse, the complete loss of IRAP expression in muscle, fat, and brain including hippocampus was accompanied with loss of binding sites for the radioligand [¹²⁵I]Nle1-AngIV. However, the IRAP-null genotype did not significantly affect the glucose transporter GLUT4 levels in the hippocampus and cortex. There was no detectable difference in the response to depolarization-evoked glucose uptake in the hippocampal slices of the IRAP-null mice compared with their wild-type littermates. Comprehensive neurologic testing of adult IRAP-null mice confirmed that the sensory perception, their motor reflexes, and vestibular function were normal. The performance of young knockout mice in memory and behavioral testing paradigms was similar to the wild-type mice. In the IRAP-null mice, the sensory, locomotor, and anxiety responses were not altered at 3–4 months of age. Furthermore, the IRAP knockout mice demonstrated improved cognitive and spatial memory when tested at 6 and 12 months of age on the Y maze task. A limited impact of IRAP-null genotype in enhancing memory was unexpected, questioning the basis of linking IRAP-inhibition to memory and cognitive enhancement.

The unexpected phenotype could be the result of developmental alteration due to germ-line deletion of the IRAP gene causing susceptibility to age-related cognitive decline. Several questions regarding IRAP-null mice remain unanswered. For instance, whether cognitive enhancement from AngIV and analogs of AngIV is completely abolished in the IRAP-null mice. The authors have demonstrated that IRAP-null mice show complete loss of AngIV binding using tissue imaging methods; they did not examine whether central infusion of AngIV or other IRAP inhibitors had any functional effect on improving performance in a range of memory paradigms or whether a compensatory mechanism is activated as a stand-in for the loss of IRAP in specific neurons. Wallis et al. (2007) reported increased levels of plasma AVP in the IRAP-null mouse, suggesting that IRAP may play a role in regulating levels of vasopressin. However, it was difficult to reconcile IRAP−/− phenotype with increased AVP and rather extensive literature on AVP effects on different brain regions. Does IRAP play a role in only disease paradigms while an unknown molecule functions as the mediator of physiologic effects of AngIV? Loyens et al. (2011) attempted to examine this question in a recent report, demonstrating that IRAP-null mice show decreased susceptibility to pentylenetetrazol (PTZ)-induced seizures. They observed that thresholds for myoclonic twitch and generalized clonic seizures significantly increased in IRAP(−/−) mice compared with their IRAP(+/+) littermates when PTZ is injected demonstrating that that IRAP(−/−) mice are less sensitive to the development of PTZ-induced seizures. The authors concluded that IRAP is involved in generalized seizure generation in normal mice. Pham et al. (2012) demonstrated that deletion of the IRAP gene protects the brain from ischemic damage analogous to the effect of the IRAP inhibitor AngIV. Transient middle cerebral artery occlusion in mice produces stroke-induced neurologic deficits 24 hours after reperfusion in mice. The infarct volumes measured were significantly reduced in the IRAP knockout mice compared with wild-type littermates, with corresponding improvement in neurologic performance. Cerebral blood flow was increased in the IRAP knockout animals.

2. Nuances for Defining Receptor(s) for AngIV

More than a decade of research focused on delineating the IRAP/AngIV binding site as the sole AngIV binding site has exposed some pharmacological, signaling, and genetic criteria not satisfactorily met by this protein. Other binding proteins have been proposed as alternative to the concept of IRAP/AngIV binding site for AngIV (De Bundel et al., 2008; Stragier et al., 2008; Wright et al., 2008).

AP-N, another transmembrane peptidase structurally related to IRAP/AngIV binding site was suggested as an additional candidate AngIV receptor, because this protein can activate IP3-dependent intracellular calcium and MAPK signaling (Navarrete Santos et al., 2000b,c). However, evidence demonstrating dimerization of AP-N and signaling in response to AngIV binding is lacking. More importantly, the role of this protein in AngIV-mediated cognitive enhancement is currently not established.

The ability of AngIV analogs to alter cognitive function and augment neurite outgrowth in vitro led to the hypothesis that the c-Met tyrosine kinase receptor for the hepatocyte growth factor (HGF) is a candidate AngIV binding site (Wright et al., 2008). Sequence-based prediction identified the possibility for AngIV and its analogs to disrupt HGF interaction with c-Met. As predicted, in picomolar concentration Norleual-AngIV was found to displace HGF binding to c-Met and antagonize signaling activity of HGF through c-Met in vitro. Norleual-AngIV blocks HGF-induced proliferation, invasion, cell scattering, and wound closure in vitro; inhibits ex vivo angiogenesis; and attenuates melanoma lung colonization in vivo. Some of the HGF/c-Met functions partially overlap with classic functions attributed to the AngIV binding site, including facilitated memory consolidation and hippocampal LTP and calcium signaling. These studies suggest that the c-Met is a cellular receptor candidate for mediating biologic effects of picomolar concentrations of AngIV analogs. These authors suggested MSP/Ron receptor system as an additional potential target of AngIV analogs that warrants investigation (Wagh et al., 2008). A number of questions regarding c-Met as a potential AngIV binding site remain unexplored. For instance, can AngIV specifically activate and AngIV-antagonists inhibit the c-Met receptor signaling in vivo to reproduce functions attributed to classic AngIV/AngIV binding site system? Does the distribution of c-Met receptor in brain and its levels within cognitive centers correlate with the classic mapped AngIV binding sites? Does LVV-H7 and the other IRAP-selective ligands discussed above bind to the c-Met receptor? Does recombinant c-Met show high-affinity binding to AngIV and LVV-H7?

3. AngIV Binding Site Conclusion.

The state of IRAP/AngIV binding site protein as the sole receptor for mediating in vivo functions of AngIV will have to be revisited in future. Multiple lines of evidences suggest that IRAP/AngIV binding site protein is a high-affinity AngIV binding receptor that mediates some of the neuroprotective and cognitive effects of AngIV, although the exact signaling mechanisms mediating these effects have not been clearly established. The IRAP/AngIV binding site hypothesis has stimulated development of rather potent and specific inhibitors of its enzymatic function, which have been shown to be cognitive enhancers in vivo. At present existing data indicate that IRAP/AngIV binding site protein does not discriminate between agonist and antagonist ligands and does not directly account for transmembrane signaling. Better characterization of IRAP/AngIV binding site inhibitors using a broad range of assays may unravel intricacies of AngIV functions in vivo. Future developments in this field are likely to uncover additional AngIV receptors and mechanisms of AngIV action in vivo. Such receptors will have to address several aspects of AngIV physiology, for example: i) in vivo efficacy at picomolar and subnanomolar concentrations at which AngIV and its analogs were found to affect changes in physiologic function; ii) the rapid onset of the physiologic effects of AngIV in blood flow and intracellular calcium signaling or neuronal potentiation experiments; iii) pharmacological effects of the "agonists" and "antagonists" on the receptor-specific effects on downstream signaling; and iv) global genetic deletion of receptor abolishes pharmacological and physiologic effects of AngIV and increases susceptibility to neurodegeneration or decline of cognitive functions. This is critical because the AngIV/AngIV binding site system is implicated in cognitive, memory, and cerebroprotective functions, as well as disease conditions like Alzheimer’s, seizures, and Parkinson’s disease. The lack of AngIV binding site agonists and antagonists penetrating blood-brain barrier has limited our understanding of importance in the etiology and treatment of memory dysfunctions associated with dementia and degenerative diseases.