Introduction

The 380-amino acid, 7-transmembrane domain G-protein-coupled receptor (GPCR) called APJ, which is most closely related to the angiotensin II (Ang II) type 1 receptor (AT1) but does not bind to Ang II, was identified through the Human Genome Project.1 In 1998, its endogenous ligand called apelin was identified.2 Apelin and APJ are highly expressed in cardiomyocytes, endothelial cells and vascular smooth muscle cells (VSMC). Apelin activates APJ, leading to intracellular signaling both locally and in an endocrine fashion.3, 4 Since its discovery, the apelin–APJ system has emerged as an important regulator of cardiovascular homeostasis and may have a critical role in the pathophysiological development of cardiovascular diseases; it has also been regarded as a new therapeutic research and drug design target.5, 6, 7, 8, 9, 10, 11 The functions of apelin–APJ appear to be complex and have yet to be defined by consensus. The accumulating evidences have indicated that the functions of apelin–APJ may be dependent on its target organs and its physiological/pathophysiological state. The experiments performed using cardiomyocytes have shown that apelin–APJ evokes contractility and promotes growth; however, in the vasculature, apelin–APJ acts as a vasodilator as part of its cardioprotective function under physiological but not pathophysiological conditions.5, 8

Recent studies have revealed that APJ and AT1 have similar expression patterns in a number of physiological and pathophysiological states, and apelin–APJ contributes to opposing effects against Ang II–AT1.10, 11 Blood pressure in AT1 and APJ double-knockout mice was significantly elevated compared with that in AT1 knockout mice.11 Chun et al. reported that apelin–APJ antagonizes Ang II–AT1 by increasing nitric oxide (NO) production via intracellular signaling in mouse models of atherosclerosis.12 However, researchers have not yet concluded that the apelin–APJ system is a counterpart of the Ang II–AT1 system. In contrast to Chun et al., Hashimoto et al. demonstrated that in mouse models of atherosclerosis, apelin–APJ aggravates oxidative stress through increasing the nicotinamide adenine dinucleotide phosphate oxidase subunits, which are major sources of superoxide anion in the cardiovascular system.13 Similarly, Liu et al. reported that apelin promotes VSMC proliferation through the phosphoinositide 3-kinase/Akt signaling pathway.14 These results indicate that apelin–APJ may have growth-promoting effects similar to those of Ang II–AT1 in the pathological development of cardiovascular diseases. Taken together, all of these results convincingly demonstrate that apelin–APJ may have dual functions. However, the specific correlation between both GPCRs’ functions still remains controversial. In the present study, we investigated the relationship among apelin, APJ, Ang II and AT1 in an in vitro experimental model. APJ alone but not apelin–APJ may be a crucial therapeutic candidate in inhibiting Ang II–AT1.

Methods

Reagents

The following reagents and antibodies were purchased: Ang II (Sigma-Aldrich, St Louis, MO, USA); anti-phosphorylated extracellular signal-regulated kinases 1/2 (anti-pERK1/2) antibody; anti-AT1 antibody; anti-angiotensin II type 2 receptor (anti-AT2) antibody; anti-APJ antibody (Santa Cruz Biotechnology, Santa Cruz, CA, USA); anti-beta-2-adrenergic receptor (anti-β2AR) antibody (Abnova, Taipei, Taiwan); anti-extracellular signal-regulated kinases 1/2 (anti-ERK1/2) antibody; anti-pAkt antibody; anti-Akt antibody (Cell Signaling Technology, Boston, MA, USA); pertussis toxin (PTX; Seikagaku, Tokyo, Japan); and the AT2-specific antagonist, PD123319 (Sigma-Aldrich). The AT1 blocker (ARB) TA-606A was a gift from the Mitsubishi Tanabe Pharmaceutical company (Osaka, Japan). We also purchased pyroglutamyl (Glp-1) apelin-13 (American Peptide Company, Vista, CA, USA) and used this as apelin.

Vector construction

The construction of the human full-length AT1, APJ, AT2 and β2AR were performed by PCR amplification using the human Multiple Tissue cDNA panel I (Clontech Laboratories, Palo Alto, CA, USA). The construction of the double-cassette vector system was previously described.15, 16, 17 Myc tag was inserted into pcDNA5/FRT/TO (Invitrogen, Carlsbad, CA, USA) to construct pcDNA5/FRT/TO-myc-His. Then, elongation factor 1α (EF1α) promoter, multiple cloning sites and bovine growth hormone site were inserted into pcDNA5/FRT/TO-myc-His. Finally, pcDNA5/FRT/TO-EF1α-V5-His-CMV-myc-His, which contained two expression cassettes, was constructed. Thus, the vectors that expressed AT1-myc-His; APJ-V5-His; AT1-myc-His plus APJ-V5-His; AT1-myc-His plus AT2-V5-His; and AT1-myc-His plus β2AR-V5-His were created.

Cell culture and transfection

Human embryonic kidney (HEK) 293 cells were grown in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum at 37 °C. Transient transfection of HEK293 cells was performed using a Fugene6 reagent (Roche Molecular Biochemicals, Indianapolis, IN, USA) as per the manufacturer's protocols. At 30 h after transfection, the cells were serum starved for 6 h before agonist stimulation. For PTX pretreatment, the cells were incubated with 200 ng ml−1 of PTX for 3 h. For PD123319 pretreatment, the cells were incubated with 10 μM of PD123319 for 1 h. The cells were pretreated with 15, 30, 60 and 120 nM of TA-606A for 90 min as previously described.18

Immunoblotting

The cells were stimulated with Ang II (100 nM); apelin (100 nM) or Ang II plus apelin at 37 °C in serum-free media after preincubation with the indicated reagents. Then, cells were lysed in an ice-cold RIPA buffer (50 mmol l−1 Tris-HCl, pH 8.0, 150 mmol l−1 NaCl, 5 mmol l−1 EDTA, 1% Nonidet P-40, 0.1% SDS and 0.5% deoxycholate) containing 100 mmol l−1 phenylmethylsulfonyl fluoride and protease inhibitor cocktail (Nacalai Tesque, Kyoto, Japan) at 4 °C and centrifuged at 20 000 g for 60 min at 4 °C. The protein concentration of the supernatant was determined using a BCA protein assay kit (Pierce, Rockford, IL, USA). Then, 4 μg of protein from each sample was subjected to SDS-polyacrylamide gel electrophoresis (10% polyacrylamide) and was transferred to a polyvinylidene fluoride membrane. The polyvinylidene fluoride membrane was blotted with a primary antibody as specified in each case. After exposure to the primary antibodies, blots were incubated with the corresponding horseradish peroxidase-conjugated secondary antibodies (Santa Cruz Biotechnology). The reactive bands were visualized with enhanced chemiluminescence substrates and analyzed using ImageJ (National Institutes of Health, Bethesda, MD, USA).

Immunoprecipitation

Cells were stimulated with Ang II (100 nM), Ang II plus 10 nM or 100 nM apelin at 37 °C in serum-free media for 2 min. The whole-cell lysate from the RIPA buffer was incubated with anti-APJ antibody or anti-rabbit immunoglobulin G antibody overnight at 4 °C for immunoprecipitation and incubated with protein G agarose (GE Healthcare, Piscataway, NJ, USA) for an additional 3 h at 4 °C, followed by a brief centrifugation on a bench-top centrifuge. The pellet was washed three times in RIPA buffer and dissolved in the SDS-polyacrylamide gel electrophoresis-loading buffer. Immunoblotting of the heterodimer was performed using an anti-AT1 antibody.

Statistical analysis

All data were expressed as a mean±s.e.m. The data were analyzed using a Student's t-test for parametric data or a two-way analysis of variance with Tukey–Kramer multiple comparison tests for comparing multiple groups. A value of P<0.05 was considered statistically significant.

Results

Relationship between apelin–APJ and Ang II–AT1 signaling

To investigate the relationship between apelin–APJ and Ang II–AT1 signaling, ERK1/2 phosphorylation was evaluated in HEK293 cells transfected with AT1 alone or dual transfected with AT1 and APJ. There was no significant difference in the expression level of AT1 in either of the transfected cell lines. In HEK293 cells dual transfected with AT1 and APJ, ERK1/2 phosphorylation was significantly reduced compared with that of AT1 alone on Ang II stimulation (P<0.05 at 2, 5 and 10 min; Figure 1a). However, apelin stimulation dramatically increased the ERK1/2 phosphorylation level beyond the level of AT1 alone (P<0.05 at 2 and 5 min; Figure 1a). To further support the idea that non-activated APJ suppresses AT1, PTX was used to create the pharmacologically non-activated APJ. Interestingly, pretreatment with PTX attenuated the ERK1/2 phosphorylation level by apelin, resulting in the same level of ERK1/2 phosphorylation in AT1 and APJ that was observed without the ligand (Figure 1a). Although immunoprecipitation of these receptors revealed that AT1 and APJ formed heterodimers, apelin stimulation never commuted their heterodimerization (Figure 1b).

Figure 1
figure 1

The relationship between apelin–APJ and AT1 signaling. (a) HEK293 cells that expressed AT1 alone or AT1 plus APJ with or without PTX pretreatment were stimulated by Ang II or Ang II plus apelin as indicated. pERK1/2, ERK1/2, AT1 and APJ were detected by western blot. (b) HEK293 cells that expressed AT1 plus APJ were stimulated by Ang II or Ang II plus apelin as indicated. pERK1/2 was detected by western blot. Immunoprecipitation was performed using an anti-APJ antibody, followed by western blot using an anti-AT1 antibody. Data represent the mean±s.e.m. from three independent experiments (*Ang II–AT1+APJ, #Ang II+apelin–AT1+APJ, P<0.05 vs. Ang II–AT1). The molecular weights are indicated in kilodaltons (kDa).

Full size image

Specificity of PTX

To determine the relationship between apelin–APJ and Ang II–AT1, several experiments were performed. Apelin evoked ERK1/2 phosphorylation via an APJ peak at 2 min and returned to the control level at 10 min (Figure 2a). In Figure 1a, the peak level of ERK1/2 phosphorylation of Ang II–AT1 plus apelin–APJ was beyond that of Ang II–AT1 because of the level of ERK1/2 phosphorylation of apelin–APJ. AT1 coexpression did not affect the ERK1/2 phosphorylation induced by apelin–APJ (Figure 2a). The ERK1/2 and Akt phosphorylation via APJ on apelin stimulation were completely inhibited by pretreatment with PTX, whereas the ERK1/2 phosphorylation through Ang II–AT1 was not attenuated by PTX (Figures 2b-d).

Figure 2
figure 2

Specificity of PTX. (a) HEK293 cells that expressed APJ alone or AT1 plus APJ were stimulated by apelin as indicated. (b, c) HEK293 cells that expressed APJ upon apelin stimulation for 2 min in the absence or presence of PTX pretreatment. (d) HEK293 cells that expressed AT1 upon Ang II stimulation for 2 min with or without PTX pretreatment. pERK1/2, ERK1/2, pAkt, Akt, AT1 and APJ were detected by western blot. Data represent the mean±s.e.m. from three independent experiments.

Full size image

Crosstalk between AT1 and other GPCRs

To confirm the specificity of APJ for Ang II–AT1 signaling, the combination of AT1 and other GPCRs was evaluated. HEK293 cells dual transfected with AT1 and β2AR were generated with the same AT1 expression level. The ERK1/2 phosphorylation level of the dual-transfected HEK293 cells was not significantly different from that of AT1 alone on Ang II stimulation (Figure 3a). Next, we evaluated the relationship between AT1 and AT2. Because AT1 and AT2 share the same ligand, the AT2-specific antagonist, PD123319, was utilized to make the pharmacologically non-activated AT2. The coexpression of non-activated AT2 impacted the ERK1/2 phosphorylation level by the action of Ang II–AT1 (Figure 3b).

Figure 3
figure 3

Crosstalk between AT1 and other GPCRs. (a) HEK293 cells that expressed AT1 alone or AT1 plus β2AR were stimulated by Ang II as indicated. (b) AT1-expressing HEK293 cells and PD123319-pretreated AT1- and AT2-coexpressing HEK293 cells were stimulated by Ang II as indicated. pERK1/2, ERK1/2, AT1, AT2 and β2AR were probed by western blot. Data represent the mean±s.e.m. from three independent experiments.

Full size image

Relationship among apelin, APJ and ARB

To reveal whether APJ may affect AT1 blocking by ARB, TA-606A was used, and the ERK1/2 phosphorylation level at 2 min after Ang II stimulation was evaluated at various concentrations of TA-606A pretreatment (Figure 4). In HEK293 cells transfected with AT1 alone, pretreatment with 30 nM of TA-606A significantly decreased the ERK1/2 phosphorylation level (50% decrease, P<0.05). Meanwhile, in the HEK293 cells dual transfected with AT1 and non-activated APJ without ligand or PTX, the ERK1/2 phosphorylation level was decreased by 60–61% of that of AT1 alone without TA-606A, and pretreatment with 30 nM TA-606A decreased the ERK1/2 phosphorylation level by 73–75%. In contrast, in the cells dual transfected with AT1 and apelin-activated APJ, pretreatment with 120 nM TA-606A only suppressed the ERK1/2 phosphorylation level by 34% compared with AT1 alone without TA-606A (Figure 4).

Figure 4
figure 4

The relationship among apelin, APJ and ARB. HEK293 cells that expressed AT1 alone or AT1 plus APJ were pretreated with a designed concentration of TA-606A with or without a combination of PTX pretreatment as indicated. Next, cells were simulated by Ang II or Ang II plus apelin for 2 min. pERK1/2, ERK1/2, AT1 and APJ were detected by western blot. Data represent the mean±s.e.m. from three independent experiments (the decreased percentage of the ERK1/2 phosphorylation level with Ang II–AT1 without TA-606A is indicated).

Full size image

Discussion

The present study demonstrated that non-activated APJ specifically suppressed Ang II–AT1 signaling. Apelin stimulation eliminated this ligand-independent function of APJ, and the Gi/o-specific inhibitor PTX elicited the pharmacologically non-activated APJ upon apelin stimulation out of relation to its heterodimerization.

Relationship between AT1 and APJ

Non-activated APJ significantly suppressed Ang II–AT1 function, whereas apelin-activated APJ attenuated this suppression. Although Chun et al. never mentioned the relevance between the non-activating and the activating APJ, they demonstrated that AT1 and APJ form receptor heterodimers in the transfected HEK293 cells.12 Most GPCRs, traditionally considered to function as monomers, actually create homodimers and heterodimers with other GPCRs. These GPCRs’ oligomerization may exist in native tissues and may have important consequences to receptor functions.19 Interestingly, apelin stimulation never altered the heterodimerization of APJ and AT1, despite changing their function; however, we also demonstrated that APJ and AT1 formed receptor heterodimers in the transfected HEK293 cells by immunoprecipitation. Accumulating evidences of the heterodimerization of GPCRs show that heterodimerization may change their receptor functions. Levoye et al. reported that the orphan GPCR, GPR50 receptor, specifically inhibits melatonin receptor type 1A function through heterodimerization independently of its ligand.20 Occasionally agonists stabilize and/or increase the level of dimers, but they can have the opposite effect.21, 22 Gines et al. reported that dopamine D1 and adenosine A1 receptors constitutively form heterodimers in the Ltk-fibroblast cells, and the dopamine D1 receptor agonist causes a decrease in the heterodimer level, leading to a cyclic adenosine monophosphate increase.23 However, we need to distinguish the relationship between AT1 and APJ from the heterodimerization of other GPCRs. Apelin stimulation may change AT1 function through APJ without its conversion to heterodimers. Furthermore, the Gi/o inhibitor PTX was used to make a functional non-activated APJ.24, 25, 26 PTX-induced, pharmacologically non-activated APJ could also suppress the Ang II–AT1 signaling, as was the case with the physiologically non-activated APJ. Although the detailed mechanisms of this relationship have been studied, this study is the first to demonstrate that the crosstalk between APJ and AT1 was determined by whether or not they received apelin stimulation independently of their heterodimerization. These findings possibly suggest that non-activated APJ, rather than apelin–APJ signaling, had a critical role in inhibiting Ang II–AT1. Various accumulating evidences clearly imply that apelin–APJ signaling has the dual functions of cardioprotection and pathologic progression. This study's results suggested that even if the apelin–APJ signaling elicits NO production via activation of endothelial NO synthase leading to cardioprotection, apelin-activated APJ loses Ang II–AT1 suppression. Taken together, these evidences indicated that apelin–APJ may have dual functions through the regulation of Ang II–AT1 (Figure 5). When NO production via apelin–APJ signaling is not enough to neutralize the non-activating APJ function that suppresses Ang II–AT1, apelin–APJ signaling may demonstrate its pathological progressive effects. Alternatively, it may behave positively, as is the case with its protective character when there is sufficient NO production.

Figure 5
figure 5

The relationship between APJ and AT1. Both receptors form the heterodimers. Non-activated APJ suppresses AT1 signaling, whereas apelin diminishes this suppression. PTX reverses apelin function.

Full size image

AT1 was shown to have interactions with several GPCRs in a functional manner.27, 28, 29, 30 Barki-Harrington et al. reported that interactions exist between AT1 and β2AR, which are not affected by ligand stimulation, and a single receptor antagonist effectively blocks downstream signaling and trafficking of both receptors simultaneously.28 In our study, the coexpression of β2AR did not affect Ang II–AT1 signaling. AT2 works as an AT1 antagonist.29, 30 However, the pharmacologically non-activated AT2 did not affect the Ang II–AT1 signaling either. Miura et al. also demonstrated that the AT1 signaling may be mainly blocked by the AT2 signaling through their negative crosstalk in phospholipase-β3 phosphorylation rather than by heterodimerization of both receptors on the cell surface.31 Collectively, the ligand-independent function of APJ in suppressing AT1 signaling may be specific.

Role of apelin–APJ in inhibiting Ang II–AT1 by ARB

ARBs, the effective AT1 uncompetitive antagonists, have been proven to be safe and well tolerated for chronic use and work as a critical component of therapy for hypertension, atherosclerosis and heart diseases.32, 33, 34 ARBs also prolong life in humans.35, 36 In the present study, the relationship among apelin, APJ and Ang II–AT1 suggested that the non-activated APJ may work as an endogenous AT1 antagonist, whereas apelin stimulation attenuated this suppression. Therefore, endogenously enhanced ARB efficiency through the non-activated APJ became a substantial pathway for consideration. The enhancing effect of the non-activated APJ on ARBs immediately converted to an abating effect by apelin stimulation. Consequently, our findings may signify that APJ alone but not apelin–APJ signaling may contribute to inhibit Ang II–AT1 by ARB. Taken together, considering the antagonizing Ang II–AT1 effect of APJ with/without apelin, the expression level of APJ and apelin in the damaged cardiovascular cells, such as the endothelial cells and VSMC, may have a potential contribution to ARB treatment in the clinical setting.

In summary, the present study has demonstrated that APJ specifically suppresses Ang II–AT1 signaling, whereas apelin stimulation inhibits this ligand-independent function. These results suggest that the crosstalk between APJ and AT1 may be distinguishable with or without apelin stimulation. Moreover, we showed that the similarities between apelin–APJ and Ang II–AT1 were also reflected in ARB function, and may contribute to ARB treatment in the clinical setting.