Positive Modulation of Angiotensin II Type 1 Receptor–Mediated Signaling by LVV–Hemorphin-7

Hemorphins are hemoglobin β-chain–derived peptides initially known for their analgesic effects via binding to the opioid receptors belonging to the family of G protein–coupled receptor (GPCR), as well as their physiological action on blood pressure. However, their molecular mechanisms in the regulation of blood pressure are not fully understood. Studies have reported an antihypertensive action via the inhibition of the angiotensin-converting enzyme, a key enzyme in the renin–angiotensin system. In this study, we hypothesized that hemorphins may also target angiotensin II (AngII) type 1 receptor (AT1R) as a key GPCR in the renin–angiotensin system. To investigate this, we examined the effects of LVV–hemorphin-7 on AT1R transiently expressed in human embryonic kidney (HEK293) cells using bioluminescence resonance energy transfer (BRET) technology for the assessment of AT1R/Gαq coupling and β-arrestin 2 recruitment. Interestingly, while LVV–hemorphin-7 alone had no significant effect on BRET signals between AT1R and Gαq or β-arrestin 2, it nicely potentiated AngII-induced BRET signals and significantly increased AngII potency. The BRET data were also correlated with AT1R downstream signaling with LVV–hemorphin-7 potentiating the canonical AngII-mediated Gq-dependent inositol phosphate pathway as well as the activation of the extracellular signal–regulated kinases (ERK1/2). Both AngII and LVV–hemorphin-7–mediated responses were fully abolished by AT1R antagonist demonstrating the targeting of the active conformation of AT1R. Our data report for the first time the targeting and the positive modulation of AT1R signaling by hemorphins, which may explain their role in the physiology and pathophysiology of both vascular and renal systems. This finding further consolidates the pharmacological targeting of GPCRs by hemorphins as previously shown for the opioid receptors in analgesia opening a new era for investigating the role of hemorphins in physiology and pathophysiology via the targeting of GPCR pharmacology and signaling.

In this study, we attempted to link the role of hemorphins in the regulation of blood pressure and RAS with its putative direct action on AngII receptors. We hypothesized that in addition to their action on ACE hemorphins may also pharmacologically target AT1R as the key GPCR in RAS. To test this, we examined the effects of LVV-hemorphin-7 on the activation of AT1R transiently expressed in HEK293 using bioluminescence resonance energy transfer (BRET) technology, which allows the real-time assessment of the functional AT1R/Gαq coupling as well as β-arrestin 2 recruitment in live cells. Moreover, we examined the functional effect of LVV-hemorphin-7 on AT1R-mediated downstream signaling pathways by measuring the cytoplasmic Gαq-dependent IP1 production and ERK1/2 phosphorylation.

Cell Culture and Transfection
HEK293 cells were maintained at 37°C, 5% CO 2 in complete medium (Dulbecco modified Eagle medium (DMEM) containing 0.3 mg/ ml glutamine, 100 IU/ml penicillin, and 100 µg/ml streptomycin) supplemented with 10% fetal calf serum (GIBCO BRL, Carlsbad, CA, USA). Transient transfections for BRET were carried out in 96-well plates and IP1 and ERK1/2 assays in six-well plates using Lipofectamine 2000 (Invitrogen). Briefly, for BRET assays in each 96-well plate, 25 ng of AT1R-Rluc was mixed with 50 ng of either Venus-Gαq or yPET-β-arrestin 2 in 25 µl of serum-free DMEM and incubated for 5 min at room temperature. The plasmid solution was then mixed with 25 µl of serum-free DMEM containing 0.5 µl of Lipofectamine 2000 and incubated for 20 min at room temperature. Cells (10 5 in 150 µl/well) resuspended in DMEM supplemented with 10% fetal calf serum (FCS) were then incubated with the final DNA-Lipofectamine 2000 mix (50 µl/well) and seeded in 96-well plates. All assays were carried out 48 h posttransfection.

Dose-Response BRET Assay
BRET technology was used for dose-response analysis as previously described (Ayoub et al., 2015;Ayoub, 2016;Ayoub et al., 2016). Cells were first washed with 50 µl/well of phosphate-buffered saline (PBS) and treated for 30 min at 37°C with 40 µL of PBS containing or not (vehicle) the indicated doses of either AngII (control), LVVhemorphin-7, or AngII in the presence of the indicated doses of LVV-hemorphin-7 (50 nM, 0.5 µM, 10 µM, or 100 µM). Following treatment, 10 µl of coelenterazine h (Promega) in PBS was added to a final concentration of 5 µM, and BRET measurements were carried out using the Tristar 2 multilabel plate reader (Berthold, Germany) allowing sequential measurements of light emission at 480 and 540 nm. For the antagonist experiments with irbesartan, cells were first pretreated with 30 µl of irbesartan (10 µM) for 15 min at 37°C before their treatment with 10 µl of LVV-hemorphin-7 (10 µM) followed by the stimulation with 10 µl of AngII (10 nM) for 30 min at 37°C, and BRET measurements were performed as described above.

Real-Time BRET Kinetic Assay
For the real-time BRET kinetics, different protocols were used. Cells were first washed with 50 µl/well of PBS and resuspended in 40 µl/ well of PBS containing coelenterazine h (5 µM), and BRET signals were measured in real time for ~5 min to determine the baseline. In the direct protocol, after the baseline, 10 µl/well of PBS containing or not (vehicle) either 10 nM of AngII (control), 10 µM of LVVhemorphin-7, or both, was added, and BRET measurements were carried out for 50 min (Figures 3A, B). In the sequential protocol, after the baseline, BRET signals were measured in real time upon two sequential treatments (T1 and then T2) as indicated in Figures 3C, D. This consists of 10 µl/well of PBS containing or not (vehicle) 10 µM of LVV-hemorphin-7 (T1) for 10 min of BRET measurements followed by 10 µl/well of PBS containing or not (vehicle) 10 nM of AngII for 25 min of BRET measurements. The opposite order (10 nM of AngII followed by 10 µM of LVV-hemorphin-7) was also performed as described in Figures 3E, F.

IP1 Accumulation Assay
Measurement of IP1 accumulation was performed in HEK293 cells expressing either AT1R-Rluc with Venus-Gαq like for BRET assays or untagged AT1R (AT1R-WT) alone using the IP-One Tb kit (Cisbio Bioassays, France) according to manufacturer's instructions. For this, cells were first pretreated or not (control) with 10 µM of LVV-hemorphin-7 for 15 min at 37°C before stimulation or not (vehicle) with the increasing doses of AngII for 30 min at 37°C. The cells were then lysed by adding the supplied assay reagents, and the assay was incubated for 1 h at room temperature. Fluorescence emission was measured at 620 and 665 nm, 50 µs after excitation at 340 nm using the Tristar 2 multilabel plate reader (Berthold, Germany).

ERK1/2 Phosphorylation
HEK293 cells expressing either AT1R-Rluc with Venus-Gαq like for BRET assays or AT1R-WT and seeded onto a six-well plate at 10 6 cells/well were first serum starved overnight and then stimulated or not with 10 nM of AngII, 10 µM of LVV-hemorphin-7, or both, at 5 or 15 min at 37°C as indicated in Figure 4B. After treatment, cells were washed in ice-cold PBS and lysed with 250 µL/well of icecold RIPA buffer (Pierce) containing protease inhibitor cocktail (Sigma) and phosphatase inhibitor tablet (Roche). Cells were then gently lysed for 1 h at 4°C, and the cell lysates were scraped from the wells, transferred into Eppendorf tubes, and centrifuged at 15,000g for 15 min at 4°C. The protein concentration in the lysate supernatant was determined using a Pierce BCA Protein Assay Kit (Thermo Fisher Scientific). An equal protein amount from each sample was mixed with Laemmli buffer (BioRad) containing 8% β-mercaptoethanol and heated at 95°C for 5 min for 10% sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) at 100 V for approximately 1 h. Proteins were then transferred onto polyvinylidene fluoride membrane (BioRad). Membranes were incubated with blocking buffer [5% skimmed milk in PBS containing 0.1% Tween 20 (PBST)] for 1 h at room temperature and washed with PBST. Then, membranes were incubated overnight at 4°C with either primary mouse monoclonal antiphospho-p44/42 (pERK1/2) (Cell Signaling) (1:2,000 dilution in TBST containing 5% skimmed milk) for the phosphorylated proteins or primary rabbit polyclonal anti-p44/42 ERK1/2 (1:1,000 dilution in TBST containing 5% bovine serum albumin) for total proteins. Membranes were then washed with PBST three times for 5 min with gentle shaking. Horseradish peroxidase-conjugated anti-immunoglobulin G was used as a secondary antibody. After further washing, immunoreactive bands were detected by ECL chemiluminescent substrate (Thermo Fisher Scientific), and chemiluminescence was detected using the LiCOR C-DiGit blot scanner. Densitometry analysis of membrane was performed using Image Studio version 5.2 software.

Data Presentation and Statistical Analysis
The BRET data given as the ratio of light emission at 540 nm over 480 nm were first converted to "ligand-induced BRET" signals by subtracting the ratio obtained from vehicle-treated cells from the same ratio obtained from AngII/LVV-hemorphin-7-treated cells. Then, the % of responses in BRET and IP1 measurements were obtained by taking as 100% the maximal AngII-induced signal in the control condition in the different assays. All kinetic and the sigmoidal dose-response curves were fitted to appropriate nonlinear regression equations using GraphPad Prism software (San Diego, CA, USA). Statistical analyses were performed with two-way ANOVA and Tukey multiplecomparisons test to determine statistical significance between the different conditions.

ND
The statistical analysis is indicative of significance compared to the conditions with AngII in the absence of 10 µM of LVV-hemorphin-7 (control). October 2019 | Volume 10 | Article 1258 Frontiers in Pharmacology | www.frontiersin.org FIGURE 2 | Positive effect of LVV-hemorphin-7 on AT1R activation revealed by dose-response BRET analysis. HEK293 cells transiently coexpressing AT1R-Rluc and either Venus-Gαq (A, C, and E) or yPET-β-arrestin 2 (B, D, and F) were used for dose-response BRET analysis as indicated in Methods. In (A and B), cells were first pretreated or not (control) for 15 min with different doses of LVV-hemorphin-7 and then stimulated 30 min with increasing doses of AngII. In (C and D), cells were first pretreated or not (control) 15 min with 10 µM of LVV-hemorphin-7 and then stimulated 30 min with increasing doses of AngII. However, in (E and F), cells were first pretreated for 15 min with increasing doses of LVV-hemorphin-7 and then stimulated 30 min with 10 nM of AngII. As a negative control, cells transiently coexpressing PAR1-Rluc and either Venus-Gαq (G) or yPET-β-arrestin 2 (H) were first pretreated or not (control) for 15 min with 10 µM of LVV-hemorphin-7 and then stimulated 30 min with increasing doses of thrombin. Data are means ± SEM of three (G and H) or five to six independent experiments performed in triplicate. ****p < 0.0001, ***p < 0.001, **p < 0.01, and not statistically significant, p > 0.05. to control cells. In order to statistically consolidate our data, we replicated further our dose-response analysis in the absence (control) or presence of 10 µM of LVV-hemorphin-7. Indeed, our second set of data confirmed the nice left shift of the AngII dose curves for BRET between AT1R-Rluc and Venus-Gαq (p < 0.001 for the whole dose curve, n = 6) ( Figure 2C) or yPET-β-arrestin 2 (p < 0.001 for the whole dose curve, n = 5) ( Figure 2D) in cells pretreated with 10 µM of LVV-hemorphin-7 compared to control cells (Table 1). This demonstrated for the first time a positive effect of LVV-hemorphin-7 on AT1R activation transiently expressed in HEK293 cells. Based on these observations, a combination of 10 nM of AngII, leading to ~25% of AT1R-mediated response, and 10 µM of LVV-hemorphin-7 was used in the different BRET kinetics and functional assays described below.
To further confirm the positive effect of LVV-hemorphin-7 on AT1R, we also performed dose-response experiments the other way around by preincubating or not cells for 15 min with increasing doses of LVV-hemorphin-7 followed by stimulation for 20 min with 10 nM of AngII. The data showed LVV-hemorphin-7 promoting a significant dose-dependent potentiation of AngII-mediated BRET increase between AT1R-Rluc and Venus-Gαq ( Figure 2E) or yPET-β-arrestin 2 ( Figure  2F) with EC 50 in the micromolar range ( Table 1).
As negative control to our data, we also tested the effect of LVV-hemorphin-7 on another unrelated GPCR, the thrombin receptor or protease-activated receptor 1 (PAR1). The data clearly showed no significant changes in the dose-response curves of thrombin-mediated BRET increase between PAR1-Rluc and either Venus-Gαq ( Figure 2G) or yPET-β-arrestin 2 ( Figure 2H).
We also performed sequential real-time kinetics where BRET signals were measured before any cell treatment for 5 min, which positions the baseline level, followed by their treatment or not with 10 µM of LVV-hemorphin-7 (treatment 1 or T1) for 10 min, and finally by cell stimulation or not with 10 nM of AngII (treatment 2 or T2) for 25 min. The kinetics clearly showed a very weak (~10%) BRET increase occurred in cells challenged with LVV-hemorphin-7 alone after 35 min of measurements. However, the AngII-induced BRET increase within AT1R-Rluc/Venus-Gαq ( Figure 3C) and AT1R-Rluc/yPET-β-arrestin 2 ( Figure 3D) pairs was strongly higher in cells previously treated with LVVhemorphin-7 compared to those treated with vehicle ( Figures  3C, D) (Table 2). Finally, BRET kinetics were also performed where cells were first stimulated with 10 nM of AngII for ~15 to 20 min to promote BRET increase between AT1R-Rluc/Venus-Gαq ( Figure 3E) and AT1R-Rluc/yPET-β-arrestin 2 ( Figure 3F) pairs before adding 10 µM of LVV-hemorphin-7. The data showed a significant BRET potentiation beyond the AngII-induced level in both cases indicating the positive effect of LVV-hemorphin-7 can also be observed even after AngII binding to AT1R. Together, the kinetic observations are consistent with the dose-response experiments shown in Figure 2 demonstrating the positive effect of LVV-hemorphin-7 on AT1R activation in HEK293 cells.

The Positive Effects of LVV-Hemorphin-7 on AT1R-Mediated Downstream Signaling Pathways
To correlate our BRET data with the downstream signaling pathways of AT1R, we examined the effects of LVV-hemorphin-7 on AT1R-mediated IP1 production as a readout for Gαq/ phospholipase C activation and ERK1/2 phosphorylation. For this, we first used HEK293 cells transiently coexpressing AT1R-Rluc and Venus-Gαq in order to use similar experimental conditions as for BRET assays. First, we tested the putative agonistic effect of LVV-hemorphin-7 on AT1R-mediated IP1 production in cells stimulated with increasing doses of either AngII as control or LVV-hemorphin-7. As shown in Figure  5A, while AngII promoted IP1 production in a dose-dependent manner and with the expected potency (Table 1), LVVhemorphin-7 had no significant effect even at 100 µM. Next, we examined the positive effect of LVV-hemorphin-7 on AngIImediated IP1 response by stimulating cells with the increasing doses of AngII in the absence or presence of pretreatment of cells with 10 µM of LVV-hemorphin-7 for 15 min. As shown in Figure 5B, like in BRET assays, LVV-hemorphin-7 promoted a significant left shift of the AngII dose-dependent curve with a stronger (~18.6-fold increase in AngII EC 50 ) effect compared to BRET data (Table 1).
To rule out any artifact in our results due to the utilization of AT1R-Rluc coexpressed with Venus-Gαq similarly to BRET assays, we also performed IP1 experiments in HEK293 cells transiently expressing the untagged AT1R (AT1R-WT). As shown in Figure 5C, AngII nicely induced IP1 production in a dose-response manner, and cotreatment of cells with 10 µM of LVV-hemorphin-7 significantly left shifted the AngII dosedependent curve with ~25-fold increase in AngII EC 50 (Table 1). This further supports the positive effect of LVV-hemorphin-7 on AT1R-mediated IP1 response.  (A and B), BRET signals were measured before (baseline) and upon cell stimulation (arrow) with either 10 nM of AngII, 10 µM of LVV-hemorphin-7, or both during 50 min. In (C and D), BRET signals were measured before any treatment (baseline) during 5 min and following treatment or not (vehicle) of cells with 10 µM of LVV-hemorphin-7 (T1) for 10 min and then additional treatment or not (vehicle) with 10 nM of AngII (T2) during 25 min. However, in (E and F), cells were first stimulated with 10 nM of AngII, and BRET signals were measured for ~20 min before an additional treatment with 10 µM of LVV-hemorphin-7 and further BRET measurements for 25 min. Data are means ± SEM of five to eight independent experiments performed in duplicate or triplicate. ****p < 0.0001, ***p < 0.001, *p < 0.05, and not statistically significant, p > 0.05. Furthermore, we examined the effect of AT1R blockade using its selective antagonist, irbesartan, similarly to BRET assays. For this, we used AT1R-WT expressing HEK293 cells pretreated or not with10 µM of irbesartan followed by treatment or not with 10 µM of LVV-hemorphin-7 combined or not with a nonsaturating dose of AngII (5 nM). Notice that 5 nM of AngII in IP1 on AT1R-WT corresponds to 10 nM used in BRET assays shown in Figure 4, and this was determined based on the dose curve obtained on AT1R-WT shown in Figure 5C. As shown in Figure  5D, both AngII-mediated IP1 production and its potentiation by LVV-hemorphin-7 were completely abolished by irbesartan treatment. These data are consistent with the BRET data shown in Figure 4 and further demonstrate the engagement of the active conformation of AT1R is such pharmacological effects of LVVhemorphin-7 on AngII-mediated IP1 response.

DISCUSSION
In this study, we report for the first time the pharmacological action of LVV-hemorphin-7 on AT1R transiently expressed in HEK293 cells with an interesting positive action on AT1R activation and signaling. This was demonstrated using in vitro assays including real-time BRET for AngII-mediated AT1R-Gαq coupling as well as β-arrestin 2 recruitment as well as their related downstream signaling pathways through the measurements of IP1 production and ERK1/2 phosphorylation. Indeed, our BRET data clearly demonstrate the positive effects of LVV-hemorphin-7 on AT1R activation and its functional coupling with Gαq and its interaction with β-arrestin 2. Both dose-response and real-time kinetics revealed a strong potentiation of AngII-induced BRET signals in the presence of LVV-hemorphin-7. Such a positive action was also observed on AngII-mediated Gαq/IP1 production and ERK1/2 phosphorylation with an interesting agonistic effect of LVV-hemorphin-7 on ERK1/2 response revealed on both AT1R-Rluc and AT1R-WT expressing cells. Thus, our data suggest a differential action of LVV-hemorphin-7 on AT1R-mediated signaling with a positive allosteric modulation (PAM)-like effect on Gαq activation and β-arrestin 2 recruitment (assessed by BRET) and an agonistic action on ERK1/2 phosphorylation. To find whether this is indicative of biased effects of LVVhemorphin-7 on AT1R or not, further investigation is required.
Our seminal finding is of significant importance to hemorphin and GPCR pharmacology and their implication in human physiology. Indeed, in addition to hemorphin's effect on opioid FIGURE 5 | Positive effect of LVV-hemorphin-7 on AT1R-mediated IP1 production. HEK293 cells transiently coexpressing AT1R-Rluc with Venus-Gαq (A and B) or AT1R-WT (C and D) were used for IP1 assays. For the dose-response experiments, cells were stimulated or not with the increasing doses of either AngII or LVV-hemorphin-7 (A) or first pretreated or not (control) with 10 µM of LVV-hemorphin-7 for 15 min and then stimulated 30 min with increasing doses of AngII (B and C) before IP1 measurements as described in Methods. In (D), cells expressing AT1R-WT were first pretreated or not (control using DMSO) with 10 µM of irbesartan for 15 min at 37ºC. Then, cells were stimulated or not (basal) with either 5 nM of AngII, 10 µM of LVV-hemorphin-7, or both, for 30 min at 37ºC before IP1 measurements were performed as described in Methods. IP1 data are means ± SEM of three to seven independent experiments performed in duplicate or triplicate. ****p < 0.0001, ***p < 0.001, and not statistically significant, p > 0.05.
receptors (Brantl et al., 1986;Davis et al., 1989;Liebmann et al., 1989;Glämsta et al., 1992;Yukhananov et al., 1994;Nyberg et al., 1997;Szikra et al., 2001;Cheng et al., 2012), we report yet another GPCR member (AT1R) that is a target of hemorphins with the plausible implication in the physiology and pathophysiology of vascular and renal systems. The action of LVV-hemorphin-7 on AT1R is consistent with the previous observations demonstrating LVV-hemorphin-7 acting on opioid receptors extending the spectra of action of hemorphins on GPCRs. For the opioid receptors, a partial to full radioligand binding was reported with a competition with some endogenous enkephalin-and dynorphinrelated opioid peptides but not nonpeptide ligands (Garreau et al., 1995;Zhao and Piot, 1997;Szikra et al., 2001). Moreover, LVVhemorphin-6 (Yukhananov et al., 1994) and-7 (Szikra et al., 2001) showed agonist-like effect as demonstrated in guinea-pig ileum and GTPγS binding assays, respectively. This clearly demonstrated the binding of these hemorphins on the opioid receptors that explain the analgesic effects of hemorphin observed in vivo (Hughes et al., 1975;Moisan et al., 1998;Sanderson et al., 1998;Albiston et al., 2004;Cejka et al., 2004;Lee et al., 2004;Cheng et al., 2012). Of course, our data do not prove a direct binding of LVV-hemorphin-7 on AT1R, but the strong positive effects observed in IP1 and BRET assays, as well as the agonistic action of LVV-hemorphin-7, observed in ERK1/2 assay and to very less extent BRET (100 µM), FIGURE 6 | Positive effect of LVV-hemorphin-7 on AT1R-mediated ERK1/2 phosphorylation. HEK293 cells transiently coexpressing AT1R-Rluc with Venus-Gαq (A and B) or AT1R-WT (C and D) were used for ERK1/2 experiments. For this, cells were stimulated 5 or 15 min with 10 nM of AngII, 10 µM of LVV-hemorphin-7, or both, as indicated, and SDS-PAGE followed by Western blot for the phosphorylated as well as total ERK1/2 was carried out and described in Methods. The representative experiment for each transfection is shown in (A and C). (B and D) represent the quantification of Western blot data for phospho-ERK1/2 changes using densitometry analysis, and the bars are means ± SEM of four independent experiments. The stars represent the statistical significance relative to the control condition (untreated cells), as well as the significance between the different treatments as indicated. ***p < 0.001, *p < 0.05, and not statistically significant, p > 0. 05. suggest such an interaction between LVV-hemorphin-7 and AT1R. Therefore, our speculation on the pharmacological effects of LVV-hemorphin-7 on AT1R emphasizes various possibilities and scenarios. First, our IP1 and BRET dose-response curves are typical of PAM of LVV-hemorphin-7 on AT1R with no significant effects of LVV-hemorphin-7 when applied alone and a nice left shift of the AngII dose curves in the presence of increasing doses of LVV-hemorphin-7. Of course, this implies that AT1R can be subjected to allosteric modulation and may present an allosteric binding site as suggested in many previous studies. For instance, an old study by Boulay et al. (1992) reported a negative allosteric modulation of AT1R by polyvinyl sulfate characterized by an inhibition of AngII affinity. Recently, PAM of AT1R by mechanical stretch of the cells has been reported (Tang et al., 2014;Wang et al., 2018). Moreover, another recent study reported a PAM of AT1R by homocysteine inducing an aggravation of the vascular injury (Li et al., 2018). Interestingly, the binding experiments showed in this study indicated the binding of homocysteine to the orthosteric binding site of AT1R along with AngII with a putative allosteric interaction between both within the binding pocket (Li et al., 2018). More recently, an allosteric AT1R-selective nanobody has been developed and used to stabilize the active conformation of AT1R and to crystalize it (Wingler et al., 2019). All these studies indicate the susceptibility of AT1R to be allosterically modulated by different kind of molecules (amino acids, nonpeptides, mechanical), and our study extends this feature to peptides (hemorphins) as potential allosteric modulators of AT1R. The other possibility to explain our data is the targeting of AT1R dimers/oligomers. In such a scenario, it is possible that no allosteric binding site exists, and only binding to the orthosteric site in AT1R occurred. This implies that both LVV-hemorphin-7 and AngII might bind in the orthosteric binding sites on two AT1R protomers, resulting in allosteric communication and effects within the AT1R dimers that lead to an increase in AngII binding and/or potency and efficacy. The putative binding of LVV-hemorphin-7 in the orthosteric binding site of AT1R would be consistent with the recent study on the allosteric modulation of AT1R by homocysteine showing binding of homocysteine to the orthosteric binding site of AT1R along with AngII with a putative allosteric interaction between both within the binding pocket (Li et al., 2018). In our study, ERK1/2 experiments clearly showed an agonistic action of LVVhemorphin-7 on AT1R similarly to AngII, suggesting the possible implication of AT1R orthosteric binding site. Thus, we cannot exclude that ERK1/2 data reflect LVV-hemorphin-7 targeting AT1R dimers. Furthermore, previous studies showed binding of LVV-hemorphin-7 on opioid receptors partially to fully compete with the binding of some endogenous enkephalin-and dynorphinrelated opioid peptides but not nonpeptide ligands (Garreau et al., 1995;Zhao and Piot, 1997;Szikra et al., 2001). Moreover, LVV-hemorphin-6 and -7 showed agonist-like effect and GTPγS binding assays (Yukhananov et al., 1994;Szikra et al., 2001). This suggests the binding of hemorphins in the orthosteric binding site of the opioid receptors without excluding the existence of another nonopioid binding site. Moreover, such a PAM of GPCRs involving effects on homodimers/heterodimers is a very important and exciting topic, and it has been reported in many occasions (Pin et al., 2005;Christopoulos, 2014;Liu et al., 2017). Finally, we cannot exclude an indirect effect of LVV-hemorphin-7 on AT1R through binding to another membrane AT1R-interacting protein expressed in HEK293 cells, resulting in transactivation of AT1R upon AngII binding. Further structural and functional studies are being considered to fully understand the exact mechanism of LVV-hemorphin-7 action on AT1R and its modulation. The profiling of other forms of hemorphin peptides should highlight further structure-function relationship data.
From the physiological point of view, the positive effects of LVV-hemorphin-7 on AT1R interestingly suggest a hypertensive action. This is somehow inconsistent with the antihypertensive effects of hemorphins reported in many in vitro and in vivo studies (Figure 7) (Lantz et al., 1991;Zhao and Piot, 1997;Fruitier-Arnaudin et al., 2002;Cejka et al., 2004;Ianzer et al., 2006;Dejouvencel et al., 2010). For instance, hemorphins have been shown to produce antihypertensive effects via the inhibition of ACE (Zhao and Piot, 1997;Fruitier-Arnaudin et al., 2002). Moreover, hemorphins have been shown to increase the hypotensive effect of bradykinin (Ianzer et al., 2006). Finally, serum hemorphin-7 levels were found to be very high after long-distance running (Glämsta et al., 1993), whereas they were very low in obese and diabetic patients, which are linked to hypertension and cardiovascular risks (Maraninchi et al., 2013). The contrasting effects of hemorphins on AT1R (activation) and ACE (inhibition) and their physiological consequences, hypertensive and hypotensive, respectively, illustrate the complexity of hemorphin's action on RAS. One way to reconcile these contrasting effects is by speculating that the inhibition of ACE that leads to a reduction of AngII levels may be counterbalanced by the positive action of hemorphins on AT1R. This would imply that, in this condition, low doses of AngII might be sufficient to activate RAS. Furthermore, one could argue that hemorphins control blood pressure and water reabsorption under various circumstances by biasing RAS toward activation through their positive effects on AT1R or inhibition of ACE activity and AngII levels. From our point of view, our data can be interpreted within both physiological and pathophysiological contexts. Indeed, it is not really clear whether the release of hemorphins from the hemoglobin constitutes a physiological or a pathogenic process with an impact on blood pressure and water balance. Moreover, it is not clear which component, hemorphins, AT1R, or ACE, is really the determinant one in the system and how this is involved in both physiological and pathophysiological situations. In the physiological situation, the dual effect of hemorphins may be part of the fine regulation of the blood pressure (and maybe water balance as well) through the positive and the negative actions on AT1R and ACE, respectively, in order to keep the blood pressure within the normal ranges. For instance, one would argue that if high blood pressure occurs for any reason and this might be associated with high expression and/or activity of ACE, hemorphins might be then released to counterbalance this situation in order to avoid any risk of sustained or chronic hypertension. In the same line, if a hypotensive/vasodilatation situation comes to occur because of low expression/activity of AT1R, hemorphins may be then released to increase blood pressure by promoting AT1R-dependent vasoconstriction. In the case of a pathological situation, the high expression/activity of AT1R associated with a plausible decrease in ACE expression/ activity and the high release of hemorphins may be symptomatic of the pathophysiology of RAS and hypertension. In that case, hypertension might be either the reason or the consequence of hemorphin's action. Thus, we believe that the system involving all the three components (hemorphins, AT1R, or ACE) may depend on the circulating concentrations of hemorphins along with the relative expression levels of the two targets, ACE and AT1R, in the different physiological and pathophysiological circumstances. So far, there are no reliable data on the plausible relationship between the concentrations of hemorphins released in blood and the expression profiles of AT1R and ACE and its impact on the control of blood pressure and hypertension. This implies that the stoichiometry hemorphins/AT1R versus hemorphins/ACE is not completely clear under hypotensive and hypertensive situations. In one old study, it has been reported that long-distance running increased the blood content in hemorphins (Glämsta et al., 1993). This may be consistent with either a positive action of hemorphins (hypertensive) through targeting AT1R because of the exercise or their negative regulatory effect (hypotensive) through the inhibition of ACE to avoid any excessive hypertension. In addition, very low hemorphin levels were recently reported in obese and diabetic patients that are linked to hypertension and cardiovascular risks suggesting beneficial effects of hemorphins on the inhibition of ACE (Maraninchi et al., 2013). Finally, it would be interesting to investigate the plausible effects of hemorphins on other key actors of RAS and kidney such as adrenergic, bradykinin, and vasopressin receptors, known to have functional interactions with AT1R (Tóth et al., 2018).

DATA AVAILABILITY STATEMENT
All datasets generated for this study are included in the manuscript.

AUTHOR CONTRIBUTIONS
AAl, AP, AAs, IA, and MA performed the research and analyzed the data; BB analyzed the data; MA and RV conceived the project, wrote the manuscript, and managed the project and its funding.

FUNDING
This work was supported by the United Arab Emirates University startup grant (31S305) to MA and a UPAR grant (31S243) to RV.