Angiotensin IV Is a Potent Agonist for Constitutive Active Human AT1 Receptors

The octapeptide hormone, angiotensin II (Ang II), exerts its major physiological effects by activating AT1 receptors. In vivo Ang II is degraded to bioactive peptides, including Ang III (angiotensin-(2–8)) and Ang IV (angiotensin-(3–8)). These peptides stimulate inositol phosphate generation in human AT1 receptor expressing CHO-K1 cells, but the potency of Ang IV is very low. Substitution of Asn111 with glycine, which is known to cause constitutive receptor activation by disrupting its interaction with the seventh transmembrane helix (TM VII), selectively increased the potency of Ang IV (900-fold) and angiotensin-(4–8), and leads to partial agonism of angiotensin-(5–8). Consistent with the need for the interaction between Arg2 of Ang II and Ang III with Asp281, substitution of this residue with alanine (D281A) decreased the peptide's potency without affecting that of Ang IV. All effects of the D281A mutation were superseded by the N111G mutation. The increased affinity of Ang IV to the N111G mutant was also demonstrated by binding studies. A model is proposed in which the Arg2-Asp281 interaction causes a conformational change in TM VII of the receptor, which, similar to the N111G mutation, eliminates the constraining intramolecular interaction between Asn111 and TM VII. The receptor adopts a more relaxed conformation, allowing the binding of the C-terminal five residues of Ang II that switches this “preactivated” receptor into the fully active conformation.

The octapeptide hormone, angiotensin II (Ang II), 1 elicits a variety of physiological and pathophysiological effects by binding and activation of angiotensin receptors belonging to the superfamily of G-protein-coupled receptors (GPCR). The majority of these effects involved in the regulation of the hydromineral balance as well as arterial blood pressure are mediated by AT 1 receptors (1). Although Ang II is considered to be the main effector of the renin-angiotensin system, other angiotensin peptides are also capable of exerting biological effects. Removal of the N-terminal amino acid (Asp 1 ) of Ang II results in the formation of Ang III. Ang III is a potent agonist of the AT 1 receptor, and was suggested to be the predominant effector peptide in the brain (2). Removal of one more amino acid (Arg 2 ) from Ang III leads to the formation of Ang IV, which binds with high affinity to the "AT 4 receptor," which is pharmacologically distinct from the AT 1 and AT 2 receptors (1,3). Although Ang IV binding sites have recently been identified as the insulin-regulated aminopeptidase (4), this finding does not explain the AT 1 receptor blocker-sensitive biological effects of Ang IV (5).
Understanding of the mechanism of AT 1 receptor and other GPCR activation has been advanced by the experimental support for the existence of multiple active or inactive conformations (6). In this context mutation-induced constitutive activation has been exploited to study the residues involved in the receptor activation process either by direct interaction with Ang II or by participating in intramolecular interaction networks that stabilize inactive receptor conformations (7-9, 10, 13). An important advance was realized by the identification of Asn 111 , located in TM III of the AT 1 receptor, to be implicated in the process of receptor activation (7,10,13). Substitution of Asn 111 by Gly was reported to cause partial ligand-independent/constitutive activation of the receptor. It has been proposed that this amino acid interacts with Tyr 292 in TM VII in the non-activated receptor and that activation of the receptor would disrupt this interaction allowing Tyr 292 to interact with Asp 74 (10). Asn 111 has also been suggested to interact with Asn 295 in TM VII (7), but other studies argue against the role of Asn 295 in stabilizing the inactive conformation of the receptor (9,11). Nevertheless, the role of Asn 111 in the stabilization of the inactive conformation of the receptor by interacting with residue(s) in TM VII is generally accepted (12).
Noda et al. (13) proposed that formation of the active receptor conformation (R*) requires two distinct Ang II-dependent steps (R-RЈ-R*). In this model, Tyr 4 of Ang II plays an important role during the initial rate-limiting step (conversion of R into RЈ) by interacting with Asn 111 and releasing it from its constraining intramolecular interaction with TM VII. Phe 8 of Ang II was also proposed to be involved during this initial step (9,13), and it is widely accepted that the N111G constitutive active receptor adopts this partially active conformation. In the second step other distinct structural features of angiotensin II cause the stabilization of the fully activated state of the receptor (R*).
The interaction between Arg 2 and Asp 281 located in TM VII of the AT 1 receptor is necessary for the high affinity binding of angiotensin peptide ligands (14). In the present study we have investigated the interaction between different angiotensin peptides and wild type, partially activated (N111G) and Asp 281 mutant (D281A) AT 1 receptors to study the role of Asp 1 and Arg 2 during receptor activation. It is shown that Arg 2 is only implied during the initial step of agonist-induced activation of the receptor and that Ang IV is a full and potent agonist for the N111G constitutive active receptor.

EXPERIMENTAL PROCEDURES
Mutagenesis of Human AT 1 Receptor DNA, Cell Culture, and Transient Transfection-The human AT 1 receptor gene, cloned in the mammalian expression vector pcDNA3 (Invitrogen), was used for expression and mutagenesis. The mutation of Asn 111 to Gly, Asp 281 to Ala, or their combination was created by the polymerase chain reaction method using the Mutagene kit (Bio-Rad). The DNA sequence of mutated receptor was confirmed by dideoxynucleotide sequencing (Amersham Biosciences). The wild type and the mutated receptors were transiently expressed in CHO-K1 cells and cultured as described previously (15).
Measurement of Inositol Phosphate Accumulation-Transfected cells were labeled for 20 h with 1 Ci/ml myo-[ 3 H]inositol (Amersham Biosciences). The labeled cells were washed twice with DMEM medium and incubated with this medium containing 10 mM LiCl for 15 min. Then agonists were added, and the incubation was continued for another 15 min at 37°C. At the end of incubation, the medium was removed, and the total soluble inositol phosphate (IP) was extracted from the cells as described previously (15). The amount of [ 3 H]IP eluted from the column was counted and a concentration-response curve generated using non-linear regression analysis. DiVal Ang IV was synthesized by Dr. G. Munske (Washington State University).
Radioligand Binding Assays-Competition binding experiments were carried out at 37°C for 1 h on intact adherent cells as described previously (16)

RESULTS
Wild type human AT 1 receptors and mutant receptors, in which Asn 111 was replaced by glycine (N111G), Asp 281 was replaced by alanine (D281A), or in which both replacements were combined, were transiently expressed in CHO-K1 cells. Their function was assessed by measuring the inositol phosphate accumulation. The higher basal level of inositol phosphates in cells with the N111G and the double mutated receptor agrees with previous findings (7,10,13). As shown in Fig 1 and Table I, Ang II and its N-terminal amino acid-deleted fragment Ang III were almost equipotent agonists for the wild type receptor. Ang IV (deletion of two N-terminal amino acids) and DiVal Ang IV were full agonists but with much weaker potency. On the other hand the C-terminal pentapeptide angiotensin-(4 -8) was barely active and the tetrapeptide angiotensin-(5-8) was completely inactive at concentrations up to 100 M. The N111G substitution only minimally affected the stimulatory properties of Ang II and Ang III, but it produced a dramatic increase in the potency of Ang IV (900-fold) and of DiVal Ang IV (1500-fold) ( Table I). Concentration-response curves of these four investigated agonists were now alike for the N111G-mutated AT 1 receptor. The potency of angiotensin-(4 -8) also increased dramatically, and it became a full agonist with a somewhat lower potency (EC 50 ϭ 28.3 nM) compared with Ang II. On the other hand, angiotensin-(5-8), which lacks the important Tyr 4 residue, was only a very weak and partial agonist for the N111G-mutated receptor (Fig. 1). In parallel, the binding affinities of Ang II, Ang III, Ang IV, and DiVal Ang IV were investigated by competition binding experiments with the highly potent and AT 1 receptor-selective antagonist  binding experiments revealed that candesartan has 100 times lower affinity for the N111G-mutated receptor as compared with the wild type receptor. In agreement with the functional data, the binding affinity of the investigated angiotensin-related peptides was increased for the N111G-substituted receptor. Whereas only a relative weak increase in affinity was seen for Ang II and Ang III, a substantial rightward shift of the competition curves occurred for Ang IV and DiVal Ang IV (Table I). Furthermore, specific binding of [ 125 I]Ang IV was only measurable in N111G-mutated but not in wild type AT 1 receptor-expressing cells (Fig. 2).
The impact of the D281A mutation on binding and function of the angiotensin related peptides was examined. The concentration-response curves for Ang IV and DiVal Ang IV were only modestly affected by the D281A substitution (Fig. 1). [ 3 H]Candesartan binding (Table I) was also only marginally affected by this mutation. On the other hand, the concentration-response curves of the Ang II and Ang III were shifted to the right by 33and 66-fold, respectively, and their potency was also markedly reduced in competition binding experiments. The effects of the D281A substitution could be completely reversed by the N111G mutation. The functional properties of the angiotensin peptides as well as their binding affinities were the same in N111G/ D281A when compared with the single N111G substitution. DISCUSSION Insight into the molecular mechanism by which Ang II activates the AT 1 receptor has been gained by structure-activity relationship studies with peptide analogs as well as receptor mutants. In this context Arg 2 as well as the aromatic residues Tyr 4 and Phe 8 of Ang II have essential roles in the binding and receptor activation (12,17,18). Measuring inositol phosphate production in human AT 1 receptor-expressing CHO-K1 cells, we show that the N111G mutation produces a dramatic increase in the potency of Ang IV and DiVal Ang IV, while only causing a moderate increase in the potency of Ang II and Ang III. In this way, all four peptides become similarly potent agonists. Moreover the C-terminal pentapeptide angiotensin-(4 -8) displays very little activity at the wild type receptor at the concentration range tested, but it becomes a full agonist for the N111G-mutated receptor. It is likely that the large increase in agonist activity of angiotensin-(4 -8) on the N111G mutant receptor was also caused by an increase in potency, since earlier studies have demonstrated that at very high concentra-tions angiotensin-(4 -8) is a full agonist of the AT 1 receptor in adrenal glomerulosa cells (20). The somewhat lower potency of angiotensin- (4 -8) to stimulate the N111G mutant receptor compared with that of Ang II, Ang III, and Ang IV may indicate that Val 3 of the ligand is auxiliary for optimal agonist binding. On the other hand angiotensin-(5-8) only became a very weak and partial agonist for the mutated receptor. In the same line, competition binding studies with the AT 1 receptor-selective antagonist [ 3 H]candesartan reveal that the N111G mutation increases the affinity of Ang IV and DiVal Ang IV, and the mutated receptor can now directly be detected by binding of [ 125 I]Ang IV. Since the conformation of the receptor in its preactivated state is mimicked by the N111G mutant, the present data suggest that Ang IV and angiotensin-(4 -8) are capable of converting the preactivated state into a fully activated conformation, but that they lack the mechanism required for the preactivation itself. The considerable lower potency and efficacy of angiotensin- (5)(6)(7)(8) to activate the preactivated state highlights the importance of Tyr 4 of Ang II in the process of AT 1 receptor activation (13). These findings are also in agreement with a previous study dealing with the aldosterone production by adrenal glomerulosa cells, which were induced by Ang IV and angiotensin-(4 -8), but not by angiotensin-(5-8) that lacks the Tyr 4 (19). The higher potency of all agonists, including angiotensin- (5)(6)(7)(8), in this study as compared with the present one could be related to a large amount of spare receptors in the adrenal glomerulosa cells (19) but not in AT 1 receptor-expressing CHO-K1 cells (16). The weak potency of Ang IV and angiotensin-(4 -8) to preactivate the wild type receptor is caused by the absence of Arg 2 , because Ang III (which is only different from Ang IV in this residue) is a fully potent agonist. On the other hand Asp 1 of Ang II is not involved in this process, since Ang III displays the same characteristics both for the wild type and the mutant receptor. This observation is in agreement with previous findings indicating that Asp 1 of Ang II is not required for agonism and only weakly interacts with amino acid residues such as His 183 of the receptor (14).
Noda et al. (13) have proposed that full activation of the AT 1 receptor is preceded by the formation of a preactivated state. It was also suggested that the N111G mutation mimics the preactivated state of the receptor, because it impedes the interaction of Asn 111 with TM VII that stabilizes the inactive conformation and, as a consequence, produces constitutive activation (13). To explain our findings and previous data we propose that Arg 2 of Ang II plays an important role in this preactivation process or, in other words, that Arg 2 participates in the destabilization of the receptor's inactive conformation. This preactivation goes along with the opening of a transmembrane pocket, which provides optimal binding of the C-terminal amino acids. Furthermore, our model shows that once the receptor is in this preactivated state, the five C-terminal amino acids contain the structural features that are necessary for full receptor activation. The present model, which is illustrated in Fig. 3, implies that while Arg 2 is necessary for the high affinity binding of peptide ligands to the wild type receptor, it is not mandatory for full receptor activation. Indeed, Ang IV is a full agonist, and its low potency suggests that receptor preactivation can occur spontaneously even in the absence of Arg 2 .
To further evaluate the role of Arg 2 of Ang II, mutant AT 1 receptors were created based on the assumption that this amino acid interacts with Asp 281 located at the extracellular end of TM VII of the receptor (14). Both our functional and binding studies on Asp 281 to alanine mutated receptor are in agreement with this proposal and previous findings. When measuring inositol phosphate production, this mutation caused a 33-fold and 66-fold increase of the EC 50 of Ang II and Ang III and, likewise, a dramatic reduction in their affinity in competition binding studies. As expected, the D281A mutation had only marginal effect on the potency of Ang IV and DiVal Ang IV. Our proposal that the Arg 2 -Asp 281 interaction is only necessary during receptor preactivation is further supported by findings with a receptor in which the N111G mutation is combined with the D281A mutation. Compared with the single D281A mutation, this double mutation rescues the impaired binding affinity and potency of Ang II and Ang III. It also produces a dramatic increase in the affinity and potency of Ang IV compared with the wild type or D281A mutant receptor. The N111G/D281A double mutant receptor not only displays the same agonist binding characteristics as the constitutively active N111G mutant, it also increases basal inositol phosphates production (data not shown), which is characteristic for the preactivated state of the receptor. Taken together, preactivation of the receptor by Ang II may involve the disruption of the Asn 111 -TM VII interaction along with the binding of Arg 2 of the peptide to Asp 281 to stabilize the new receptor conformation. Because the Asn 111 -TM VII interaction is already disrupted by the N111G mutation, this receptor mutant does not require Arg 2 to adopt the preactivated state. Asn 111 is likely to play a dual role. It stabilizes the inactive conformation of the receptor by interacting with TM VII, and it also interacts with Tyr 4 of the ligand during the process of receptor preactivation. When the receptor is preactivated this latter interaction is probably no longer required to obtain full receptor activation, since Noda et al. (13)  Altered antagonist binding is characteristic of constitutive active AT 1 receptors (7,9,10,20). It was shown previously that the N111G mutant receptor has decreased affinity to losartan, a surmountable non-peptide antagonist of the AT 1 receptor (9). Our data show that compared with the wild type AT 1 receptor the N111G mutant receptor also has 100-fold lower affinity for the insurmountable non-peptide antagonist candesartan. The reduced affinity of non-peptide antagonists to this mutant receptor suggests that, while receptor preactivation causes optimal alignment of the binding pocket for the C-terminal residues of Ang II, it also causes misalignment of the residues involved in the binding of non-peptide AT 1 receptor blockers. Since non-peptide AT 1 receptor blockers have much higher affinity to the inactive conformation of the receptor, they could act as inverse agonists irrespectively of their surmountable/ insurmountable character.
In conclusion, our data show that the two N-terminal amino acid residues of Ang II are no longer required for its high affinity binding to the N111G mutated AT 1 receptor (i.e. the receptor in a preactivated state) and that Val 3 plays only a minor role in this process. Based on these findings, we propose a model of AT 1 receptor activation that pinpoints the role of Arg 2 of Ang II in the preactivation process, i.e. it stabilizes a conformation of the receptor that is optimal for binding of the C-terminal five residues of Ang II (Fig. 3). These C-terminal residues are then sufficient to produce full receptor activation. Furthermore, whereas Asn 111 stabilizes the inactive conformation of the receptor by interacting with TM VII, it is also likely that it interacts with Tyr 4 of the ligand during the preactivation process of the receptor (13). FIG. 3. Visual representation of the experimental data of Ang II-and Ang IV-induced receptor activation. In the basal state interaction of Asn 111 in TM III with TM VII stabilizes the inactive conformation of the AT 1 receptor (left). The second amino acid (Arg 2 ) of the octapeptide hormone Ang II interacts with Asp 281 in TM VII. A conformational rearrangement or movement of TM VII disrupts the interaction between this helix and Asn 111 of TM III (upper middle). The receptor "relaxes" into a preactivated state (RЈ), and the transmembrane binding site opens to accomodate the C-terminal part of the octapeptide (upper right). The hexapeptide Ang IV lacks Arg 2 and cannot bind with high affinity to the inactive conformation of the wild type receptor (left part). In the N111G mutant receptor, the interaction between Asn 111 and TM VII is disrupted, and the receptor is relaxed into the preactivated state (RЈ) in the absence of the agonist (lower middle). Both Ang II and Ang IV can now bind with similar high affinity and cause full activation of the N111G mutant receptor (R*) (lower right).