We have previously shown that individual β-amino acid substitution in angiotensin (Ang) II reduced Ang II type 1 receptor (AT1R) but not Ang II type 2 receptor (AT2R)-binding and that the heptapeptide Ang III exhibited greater AT2R:AT1R selectivity than Ang II. Therefore, we hypothesized that β-amino-acid-substituted Ang III peptide analogues would yield highly selective AT2R ligands, which we have tested in binding and functional vascular assays. In competition binding experiments using either AT1R- or AT2R-transfected human embryonic kidney (HEK)-293 cells, novel β-substituted Ang III analogues lacked appreciable AT1R affinity, whereas most compounds could fully displace 125I-Sar1Ile8 Ang II from AT2R. The rank order of affinity at AT2R was CGP42112 > Ang III > β-Pro7 Ang III=Ang II > β-Tyr4 Ang III ≥ PD123319 >> β-Phe8 Ang III >> β Arg2 Ang III=β-Val3 Ang III >> β-Ile5 Ang III. The novel analogue β-Pro7 Ang III was the most selective AT2R ligand tested, which was >20 000-fold more selective for AT2R than AT1R. IC50 values at AT2R from binding studies correlated with maximum vasorelaxation in mouse aortic rings. Given that β-Pro7 Ang III was an AT2R agonist, we compared β-Pro7 Ang III and native Ang III for their ability to reduce blood pressure in separate groups of conscious spontaneously hypertensive rats. Whereas Ang III alone increased mean arterial pressure (MAP), β-Pro7 Ang III had no effect. During low-level AT1R blockade, both Ang III and β-Pro7 Ang III, but not Ang II, lowered MAP (by ∼30 mmHg) at equimolar infusions (150 pmol/kg/min for 4 h) and these depressor effects were abolished by the co-administration of the AT2R antagonist PD123319. Thus, β-Pro7 Ang III has remarkable AT2R selectivity determined in binding and functional studies and will be a valuable research tool for insight into AT2R function and for future drug development.
There are relatively few AT2R agonists available, although these research tools are required to probe AT2R function, particularly in chronic cardiovascular disease settings.
In the present study, we describe the binding and acute cardiovascular profile of the novel compound β-Pro7Ang III. This compound exhibited >20 000-fold AT2R selectivity over AT1R and behaved as an AT2R agonist as it evoked vasorelaxation in vitro and lowered blood pressure acutely during low-level AT1R blockade in conscious spontaneously hypertensive rats. These vascular effects were abolished by AT2R blockade.
This strategy of modifying Ang III could be used to investigate a number of related peptides and allow more detailed comparison with other AT2R agonists on cardiovascular function in chronic disease, thus providing greater information for class effects of AT2R agonists and for potential drug development using these as lead compounds.
The role of the angiotensin (Ang) II type 2 receptor (AT2R) in cardiovascular regulation is the subject of much debate. Most evidence points towards AT2R vasodilator activity and counter-regulatory mechanisms that oppose Ang II type 1 receptor (AT1R) stimulation [1–4], although precise understanding of AT2R function has been hindered by the dearth of selective compounds with which to probe function. Indeed, currently there are only two commercially available compounds: agonist CGP42112 and antagonist PD123319. In addition, there is the more previously developed non-peptide agonist compound 21 (C21), which shows good AT2R selectivity although off-target effects have been reported , albeit at concentrations 1000-fold greater than those used conventionally to stimulate AT2R. Thus it has been difficult to determine class effects of AT2R stimulation, which are necessary to confirm the AT2R as a viable therapeutic target.
In this context, we have previously reported that a subtle chemical substitution within the peptide sequence of the endogenous ligand, Ang II, results in a profound shift in AT2R:AT1R binding . Indeed, in contrast with Ang II which shows similar affinity for both receptor subtypes, β-Ile5 Ang II exhibited >1000-fold selectivity for AT2R over AT1R binding .
Although Ang II is widely acknowledged as the major bioactive peptide of the renin-angiotensin system, acting predominantly via AT1R, other Ang peptides are known to exert distinct biological effects. For example, the C-terminal heptapeptide Ang III exerts many effects that are similar to those evoked by Ang II except that this peptide is less potent at AT1R . Interestingly, Scheuer and Perrone  reported that Ang III produced a biphasic response consisting of an initial pressor response which was sensitive to AT1R blockade, followed by a depressor response which was blocked by the AT2R antagonist PD123319 . In this context, Ang III has been reported to show some selectivity over Ang II for AT2R in binding studies [9–12], and in signal transduction studies in which AT2R were transfected into COS-7 cells, Ang III could readily stimulate AT2R . Carey and colleagues [14–16] have reported that Ang III, not Ang II, is the preferred endogenous AT2R ligand, at least in terms of the kidney's natriuretic response.
Thus, given that (i) β-substitution of Ang II was able to reduce AT1R without diminishing AT2R binding, and (ii) native Ang III shows greater AT2R:AT1R selectivity than Ang II, we reasoned that a similar approach of β-substitution in the Ang III native peptide may yield AT2R-selective ligands with even higher affinity than previously synthesized. Therefore, in the present study, we have synthesized a series of β-substituted Ang III peptide analogues, such that the corresponding β-substituted amino acid has replaced the natural α-amino acid at each position in the Ang III sequence. We hypothesized that this series may exhibit enhanced AT2R:AT1R selectivity and efficacy in functional (in vitro and in vivo) vascular assays.
MATERIALS AND METHODS
β-Peptide preparation, synthesis, purification and HPLC traces, as well as all reagents sources, is described in detail in the Supplementary Online Data (Supplementary Table S1; Supplementary Figure S1).
Male 10-week-old FVB/N mice (Monash Animal Research Platform) and ∼16–18-week-old-male spontaneously hyptertensive rats (SHR) (Animal Resources Centre) were used. Animals were maintained on a 12-h light/12-h dark cycle and were fed standard laboratory mice chow and received water ad libitum. Experiments were approved by the Monash University Animal Research Platform Animal Ethics Committee and were performed in accordance with the Australian Code of Practice for the Care and Use of Animals for Scientific Purposes.
Human embryonic kidney (HEK)-293 cells stably transfected with either AT1R or AT2R were used for binding studies, as described previously . Cells were grown to approximately 80% confluence before being re-plated into 48-well plates at 1×105 cells/well and grown for 48 h at 37°C for a whole cell competition binding assay using 125I-Sar1Ile8Ang II at 50 000 cpm. The unlabelled ligands used in the competition assays were native Ang III, β-substituted Ang III peptides, CGP42112, PD123319 and candesartan at concentrations ranging from 1 pM to 1 μM. For each experiment, each ligand concentration was tested in triplicate. Non-specific binding (NSB) was defined in the presence of the unlabelled Ang II (1 μM). The ability of each ligand to inhibit specific binding of 125I-Sar1Ile8Ang II was assessed. All counts were corrected for NSB and the percentage of bound AT1R or AT2R was calculated by normalizing the specific binding at different concentrations of the competing drugs to the specific binding determined in the presence of vehicle. Non-linear regression of the data was performed and IC50 values were estimated. In cases where there was minimal displacement of 125I-Sar1Ile8Ang II by the unlabelled peptide ligand (at 1 μM), an arbitrary value of 10 μM was used as a comparative estimate of IC50 values (to be used in ratio calculations).
In vitro reactivity
Mouse thoracic aortae were removed (after isoflurane inhalation followed by decapitation) and cut into ring segments for isometric tension recording. Following equilibration for 90 min at 0.5 g resting tension, vessels were pre-contracted to examine vasorelaxation, as described previously . Tissues were pre-contracted with U46619 to attain 40–50% of the maximum contractile response and the ability of β-amino-acid-substituted Ang III peptides to evoke relaxation was determined and referenced to effects of the standard AT2R agonist CGP42112. A parallel tissue served as a time control in which only U46619 was given. At the end of the experiment, 10 μM of the endothelium-independent vasodilator sodium nitroprusside (SNP) was added to the organ bath to test the integrity of the vascular smooth muscle cells. In additional experiments, vasorelaxation in response to β-Pro7Ang III was examined in pre-contracted aortae in the absence and presence of either the AT2R antagonist PD123319 (1 μM) or the Mas receptor (MasR) antagonist A-779 (1 μM), as well as in AT2R-deficient mice.
In vivo blood pressure studies in SHR
On the day prior to experimentation, rats were anaesthetized [ketamine (75 mg/kg)/xylazine (10 mg/kg) intraperitoneal (IP), supplemented as required] and catheters were inserted into the right carotid artery and right jugular vein for direct blood pressure measurement and drug administration respectively. Approximately 24 h after surgery, the carotid artery was connected to a pressure transducer (Gould) attached to a MacLab-8 data acquisition system (ADInstruments). Mean arterial pressure (MAP) was derived from the phasic blood pressure signal. Experiments were performed in conscious SHR that underwent daily experimental protocols over 4–6 days that were similar in design to those reported previously [17,18].
Initially, separate groups of SHR were used to confirm an enhanced AT2R-mediated depressor effect of Ang III compared with that of Ang II, which was achieved by the following treatments which were randomized on separate days within the same animals: (i) infusion of saline (0.35 ml/h) for 4 h; (ii) intravenous (IV) bolus of candesartan (0.01 mg/kg) alone plus saline infusion; (iii) Ang II or Ang III infusion at 150 pmol/kg/min over 4 h; or (iv) Ang II or Ang III infusion plus candesartan (0.01 mg/kg, IV bolus). Additional groups of SHR received Ang peptide/candesartan combinations alone or in the presence of PD123319 (50 μg/kg/min; AT2R antagonist), icatabant [HOE-140; 100 μg/kg; BK (bradykinin) type 2 receptor (B2R) antagonist] or NG-nitro-L-arginine methyl ester (L-NAME) [1 mg/kg; nitric oxide synthase (NOS inhibitor)].
In a further two groups of SHR, initial dose-ranging experiments were performed with the most highly selective AT2R ligand synthesized in the present study, β-Pro7Ang III. Thus animals were infused with β-Pro7Ang III at two doses following a similar protocol to that described above. Specifically, β-Pro7Ang III infusions at 50 or 150 pmol/kg/min were performed over 4 h; either alone or in the presence of candesartan (0.01 mg/kg, IV bolus) or PD123319 (50 μg/kg/min for 4 h). A lower infusion of β-Pro7Ang III (15 pmol/kg/min) did not affect MAP.
The stability of native Ang III or β-Pro7Ang III (at 1 mg/ml) in rat plasma (diluted 4:1 with saline) incubated at 37°C was determined as described previously . Protease activity was quenched in spiked samples by the addition of acetonitrile (sample/acetonitrile, 1:1) at selected intervals and stored at −20°C for later analysis. Samples were centrifuged at 9300 g for 5 min and the supernatant was collected for LC–MS analysis. The amount of parent compound remaining at each time point was then assayed on an Agilent 1100 MSD SL ion trap mass spectrometer. The peaks observed in the resulting chromatograms were integrated, compared with a standard curve and cross-checked by mass and retention time.
All data are presented as means±S.E.M. unless otherwise stated. IC50 values, representing the concentration at which each ligand displaced 50% binding of 125I-Sar1Ile8Ang II in either AT1R- or AT2R-transfected HEK-293 cells, were estimated following non-linear regression using GraphPad Prism 6.0. Ratios of IC50 values for each ligand at AT1R relative to IC50 values at AT2R were determined as a measure of AT2R selectivity. Differences in MAP from baseline measurements were analysed (GraphPad Prism) using a one-way ANOVA with repeated measures. Differences in MAP between treatments and treatment/time interactions were analysed using a two-way ANOVA with repeated measures. Statistical significance was accepted at P<0.05.
Competition for 125I-Sar1Ile8Ang II binding in HEK-293 cells transfected with AT1R was observed for Ang II, Ang III and candesartan, with IC50 values in the nanomolar range, whereas all other ligands lacked affinity for the AT1R (IC50 values > 1μM; Figure 1A; Table 1). By contrast, in HEK-293 cells transfected with AT2R, competitive binding for 125I-Sar1Ile8Ang II was observed with the majority of ligands except candesartan and β-Ile5 Ang III (Figure 1B; Table 1). As expected, IC50 values were in the nanomolar range for Ang II, Ang III and CGP42112 and this was also evident for β-Pro7Ang III. Full displacement of 125I-Sar1Ile8Ang II binding was also noted for PD123319, β-Tyr4Ang III and β-Phe8Ang III, whereas β-Arg2Ang III and β-Val3Ang II were markedly less effective (at 1 μM). The rank order of affinity at AT2R was CGP42112 > Ang III > β-Pro7Ang III=Ang II > β-Tyr4Ang III ≥ PD123319 >> β-Phe8 Ang III >> β-Arg2Ang III=β-Val3Ang III (Table 1). With respect to AT1R binding, only candesartan, Ang II and Ang III fully displaced 125I-Sar1Ile8Ang II binding at the concentrations used, whereas β-Ile5Ang III lacked affinity for both the AT1R and the AT2R, up to at least 1 μM (Figure 1).
AT1R- and AT2R-binding of β-substituted Ang III ligands
|.||AT1R .||AT2R .||.|
|Ligand .||IC50 value (M) .||IC50 value (M) .||AT2R (fold selectivity)* .|
|β-Pro7Ang III||–†||4.68×10−10||21 377|
|.||AT1R .||AT2R .||.|
|Ligand .||IC50 value (M) .||IC50 value (M) .||AT2R (fold selectivity)* .|
|β-Pro7Ang III||–†||4.68×10−10||21 377|
*AT1R IC50/AT2R IC50.
†IC50 value of 1.00×10−5 M was used for the ratio calculation.
In order to estimate relative AT2R selectivity, the ratio of IC50 values at AT1R to that at AT2R was determined. As expected, known AT2R ligands exhibited marked selectivity, ranging from approximately 3000- to 18 000-fold AT2R selectivity for PD123319 and CGP42112 respectively, whereas candesartan was approximately 15 000-fold selective for AT1R (Table 1). Given that β-Pro7Ang III exhibited similar AT2R affinity, but less AT1R affinity, relative to CGP42112, this compound exhibited the greatest AT2R selectivity that being approximately 21 000-fold AT2R selectivity over AT1R (Table 1).
All peptides except β-Ile5Ang III were tested for AT2R agonist activity in pre-contracted mouse aortae and compared with the known AT2R agonist CGP42112. β-Pro7Ang III and β-Tyr4Ang evoked vasorelaxation of a similar magnitude to that with CGP42112 (Figure 2). Pre-incubation with the AT1R antagonist, candesartan, tended to increase vasorelaxation for β-Arg2Ang III and β-Val3Ang III. However, these responses were still less than that of CGP42112, whereas candesartan had no additional effect on β-Phe8Ang III. The AT2R antagonist PD123319 abolished vasorelaxation evoked by all β-substituted Ang III peptides. Moreover, maximum vasorelaxation of pre-contracted mouse aortae was highly correlated with AT2R affinity (Figure 2). The AT2R selectivity of β-Pro7Ang III was also confirmed by the finding that, in parallel tissues, the MasR antagonist A-779 failed to inhibit the vasorelaxation caused by β-Pro7Ang III, whereas PD123319 abolished this effect. Moreover, β-Pro7Ang III did not evoke vasorelaxation in pre-contracted mouse aortae taken from AT2R-deficient mice (Figure 3).
AT2R-mediated vasorelaxation evoked by β-substituted Ang III ligands
Functional AT2R selectivity of β-Pro7 Ang III
Acute infusions were also performed in conscious SHR. In SHR, infusion of saline alone had no effect on MAP and the AT1R antagonist candesartan was used at a dose that caused only small reductions in MAP (<15 mmHg) in the different groups of SHR, as we have reported previously [17,18]. Ang III infusion (150 pmol/kg/min) caused a significant and sustained increase in MAP (10–20 mmHg increases; Figure 4). Ang II, infused alone at the same dose, also caused a significant increase in MAP (∼40 mmHg increases; Figure 4). Strikingly, when Ang III was infused during AT1R blockade, the Ang III/candesartan combination caused a greater reduction in MAP (25–35 mmHg decreases) than candesartan alone. By contrast, in analogous experiments, the effect of Ang II co-administered with candesartan was not greater than the effect of the AT1R antagonist alone (Figure 4).
Vasopressor and/or vasodepressor effects of Ang II and Ang III
AT2R selectivity of the depressor response to Ang III was confirmed in a separate group of rats in which Ang III was infused in the presence of pharmacological blockade of signalling pathways associated with AT2R-mediated vasorelaxation. As expected, the Ang III-induced depressor effect was completely abolished by co-infusion of PD123319, HOE-140 or L-NAME (Supplementary Figures S2 and S3), thus confirming involvement of nitric oxide (NO) and BK in AT2R-mediated vasorelaxation.
Analogous experiments to those performed with Ang II and Ang III were used to assess the effect of the β-substituted Ang peptide with highest selectivity for AT2R:AT1R, i.e. β-Pro7Ang III. In direct contrast with Ang II or Ang III infusion, β-Pro7Ang III did not increase MAP when administered alone (Figure 5). However, when given against a background of low-level AT1R blockade, at an equivalent molar infusion rate, β-Pro7Ang III (150 pmol/kg/min) significantly reduced blood pressure in a PD123319-sensitive manner, confirming AT2R-mediated vasorelaxation. Importantly, this depressor response to β-Pro7Ang III was also achieved when the infusion dose was reduced by one-third of the original dose (Figure 5). The apparent half-lives of native Ang III and β-Pro7Ang III were 32 and 56 min respectively, derived from the analysis of one-phase exponential decay of the peptide in a spiked plasma sample.
AT2R-mediated vasodepressor effects of β-Pro7 Ang III
The main findings of the present study were that a β-amino acid screen of Ang III was well tolerated with respect to AT2R, but not AT1R, binding and a β-amino acid substitution at the Pro7 residue of Ang III resulted in a ligand with one of the highest AT2R to AT1R selectivities yet reported. Moreover, this AT2R-selective ligand, β-Pro7Ang III, was an agonist since it caused both in vitro relaxation and in vivo depressor effects in conscious SHR.
Ang III is widely reported to mediate similar cardiovascular effects to its precursor peptide Ang II, including vasoconstriction via the predominant AT1R, albeit at reduced potency [1,7]. Indeed, both Ang III and Ang II, when infused alone in the present study, evoked a similar increase in MAP in conscious SHR. However, the present and previous binding data support the view that Ang III displays slightly greater AT2R:AT1R selectivity than Ang II [9,12].
Results from radioligand competition binding studies indicated that β-substitutions in the amino acid sequence of Ang III differentially altered the affinity of peptides for AT1R and AT2R. As expected, native Ang III displaced binding in both AT1R- and AT2R-transfected HEK-293 cells, with IC50 values for this reference ligand, as well as for Ang II, CGP42112, PD123319 and candesartan, in excellent agreement with our previous study . We noted that β-substituted Ang III heptapeptides displayed little affinity for the AT1R, whereas several analogues, including β-Pro7Ang III and β-Tyr4Ang III, exhibited high AT2R affinity. β-Ile5Ang III exhibited minimal AT1R or AT2R binding (up to 1 μM), which was unexpected and illustrates that subtle modification can exert marked changes in ligand affinity [6,10,12]. In this instance, the removal of aspartic acid in β-Ile5 Ang III compared with β-Ile5 Ang II results in a subtle alteration in 3D structure which prevents interaction of the ligand with the AT2R-binding pocket. When IC50 binding ratios (AT2R:AT1R) were used to determine rank order of AT2R-selectivity of Ang III peptides (β-Pro7 > β-Tyr4 > β-Phe8 >> β-Arg2 > β-Val3 >>> β-Ile5), it was clear that β-Pro7 Ang III was highly selective for the AT2R over AT1R (∼21 000-fold) and comparable with the AT2R-selective agonist CGP42112, although CGP42112 itself exhibits AT1R binding at high concentrations. Interestingly, β-Pro7 Ang III was at least 10-fold more potent at AT2R and approximately 10-fold more AT2R-selective than the best functionally active AT2R agonists derived from our previous β- substituted scan using Ang II as the template .
In vascular assays, all β-substituted Ang III peptides tested caused concentration-dependent vasorelaxation, of varying degrees, with β-Pro7Ang III and β-Tyr4Ang III both evoking a similar magnitude of vasorelaxation to that of CGP42112 assessed in parallel tissues. Moreover, the maximum relaxation of each β-substituted Ang III heptapeptide correlated with individual IC50 binding values, indicating the importance of AT2R affinity for compounds to evoke maximal aortic vasorelaxation, which was consistent with the AT2R antagonist PD123319 abolishing these effects. Moreover, the findings that β-Pro7 Ang III-induced vasorelaxation was absent from AT2R-deficient mice, but preserved in the presence of the MasR antagonist A-779, confirms the specificity of this ligand as an AT2R-selective agonist.
Using our standard in vivo blood pressure (BP) assay that has identified AT2R-mediated vasodilator effects of CGP42112, C21 and β-Ile5Ang II [6,17,18], we tested the effects of native Ang III in conscious SHR. Indeed, these results indicated that Ang III exerts AT1R-mediated pressor effects, when given alone, that were converted into AT2R-mediated depressor effects in the presence of AT1R blockade. The requirement for low-level AT1R inhibition in order to demonstrate a depressor effect of AT2R stimulation is entirely consistent with previous studies [6,17,18] and is presumably due to tonic AT1R-mediated vasoconstriction which predominates over AT2R-mediated vasorelaxation due to endogenous Ang in vivo  (Figure 2). Importantly, the effect of Ang III could not be mimicked by Ang II under the same experimental conditions. These differential effects of Ang III and Ang II are in agreement with previous in vivo vascular  and sodium excretion  studies that also failed to unmask AT2R-mediated effects of Ang II. Moreover, we found that the AT2R-mediated vasodepressor effect evoked by Ang III (in the presence of candesartan) was abolished in the presence of either the B2R antagonist HOE-140 or the NOS inhibitor L-NAME. Thus, these results are entirely consistent with previous functional [19,20] and signalling [19,21,22] studies in which AT2R stimulation resulted in BK and cGMP production in conscious rats and are generally supported by a larger body of literature identifying a link between the activation of the AT2R and the production of BK and NO [3,4].
Given that β-Pro7 Ang III exhibited the highest AT2R:AT1R selectivity and evoked AT2R-mediated in vitro vasorelaxation of a similar magnitude to the standard CGP42112, we compared the in vivo effects of this analogue with Ang III. Unlike Ang III, β-Pro7Ang III itself did not evoke an AT1R-mediated pressor effect, consistent with CGP42112 and β-Ile5Ang II performed in the same model [6,18]. However, in the presence of low-dose candesartan, β-Pro7Ang III, given at an equimolar dose to Ang III, evoked a substantial decrease in MAP that was still evident 1 h after the infusion had stopped and when also tested in a separate group of SHR at one-third of the dose. These depressor effects were abolished by concomitant infusion of PD123319, confirming an AT2R involvement. In an earlier study , we noted vasodepressor responses in response to β-Ile5Ang II at 10-fold lower infusions (15 pmol/kg/min), which were consistent with a 10-fold increase in half-life of the β-substituted peptide compared with the parent peptide (Ang II). In the present study, β-Pro7Ang III only doubled the half-life of Ang III; thus this modest increase in peptide stability, relative to previous studies , may explain the lack of efficacy of β-Pro7Ang III at lower doses.
Our relatively simple strategy of incorporating β-amino acid substitutions on an Ang template [6,23] has increased AT2R:AT1R selectivity from ∼1000-fold using Ang II analogues (e.g. β-Ile5Ang II; ) to >20 000-fold with Ang III analogues such as β-Pro7Ang III. The present data also confirm the importance of Asp1 for AT1R binding since Ang III modifications (to amino acid residues 2–8) were generally well tolerated at AT2R but devoid of AT1R binding at concentrations used in the present study. Moreover, these data fit well with the concept that substitutions at the middle/C-terminal portion of Ang II/III confers marked AT2R-binding selectivity [10,24], as also observed recently when tyrosine was substituted for histidine in position 6 of Ang II . We did not study β-His6Ang III since β-histidine is not commercially available. Moreover, in the present study, we showed that AT2R-binding activity for β-substituted Ang III analogues translated into functional activity and that our lead compound, β-Pro7Ang III, lowered MAP acutely during AT1R blockade. Whereas chronic AT2R stimulation is unlikely to involve an anti-hypertensive effect, it does cause marked tissue remodelling, as exemplified by the non-peptide agonist C21 in a range of pre-clinical disease settings [26–29]. The fact that C21 also lowers MAP acutely  in a similar manner to β-Pro7Ang III suggests that it will be of interest to determine the chronic effects of our new highly selective AT2R agonist β-Pro7Ang III. Such studies will be important for comparison with C21 to determine ‘class effects’ of AT2R agonists and provide new research tools for further drug development.
Mark Del Borgo synthesized β-substituted Ang III peptides, using a chemical strategy developed initially by Marie-Isabel Aguilar and Patrick Perlmutter and performed stability assays. Sanja Bosnyak performed binding experiments. Morimer Khan and Iresha Spizzo conducted in vitro vascular experiments, whereas Yang Wang, Pia Walters and Lucinda Hilliard conducted in vivo experiments. Sanja Bosnyak, Yang Wang and Emma Jones analysed the data, whereas Mark Del Borgo and Yan Wang managed the project on a daily basis under the supervision of Emma Jones and Robert Widdop. The experiments were conceived by Robert Widdop, Emma Jones and Marie-Isabel Aguilar, and Kate Denton also provided intellectual input. The main contributors to writing the manuscript were Mark Del Borgo, Yang Wang, Marie-Isabel Aguilar, Kate Denton, Emma Jones and Robert Widdop.
We thank Professor Walter Thomas for the AT1R- and AT2R-transfected HEK-293 cells.
This work was supported by the National Health and Medical Research Council of Australia [grant number APP1045848]; and the Heart Foundation of Australia [grant numbers G09M4521 and G11M5797].
angiotensin II type 1 receptor
angiotensin II type 2 receptor
bradykinin 2 receptor
human embryonic kidney
NG-nitro-L-arginine methyl ester
mean arterial pressure
nitric oxide synthase
spontaneously hypertensive rat(s)
These authors contributed equally to the article.