Heterologous down-regulation of angiotensin type 1 receptors by purinergic P2Y₂ receptor stimulation through S-nitrosylation of NF-κB

Abstract

Cross-talk between G protein-coupled receptor (GPCR) signaling pathways serves to fine tune cellular responsiveness by neurohumoral factors. Accumulating evidence has implicated nitric oxide (NO)-based signaling downstream of GPCRs, but the molecular details are unknown. Here, we show that adenosine triphosphate (ATP) decreases angiotensin type 1 receptor (AT₁R) density through NO-mediated S-nitrosylation of nuclear factor κB (NF-κB) in rat cardiac fibroblasts. Stimulation of purinergic P2Y₂ receptor by ATP increased expression of inducible NO synthase (iNOS) through activation of nuclear factor of activated T cells, NFATc1 and NFATc3. The ATP-induced iNOS interacted with p65 subunit of NF-κB in the cytosol through flavin-binding domain, which was indispensable for the locally generated NO-mediated S-nitrosylation of p65 at Cys38. β-Arrestins anchored the formation of p65/IκBα/β-arrestins/iNOS quaternary complex. The S-nitrosylated p65 resulted in decreases in NF-κB transcriptional activity and AT₁R density. In pressure-overloaded mouse hearts, ATP released from cardiomyocytes led to decrease in AT₁R density through iNOS-mediated S-nitrosylation of p65. These results show a unique regulatory mechanism of heterologous regulation of GPCRs in which cysteine modification of transcriptional factor rather than protein phosphorylation plays essential roles.

G protein-coupled receptors (GPCRs) are the largest family of cell-surface receptors, which play a critical role in regulating multiple physiological functions (1, 2). Abnormal activation or up-regulation of GPCRs is a major cause of various diseases (3), and about 40% of drugs that are widely used for therapeutic treatment all over the world may directly or indirectly target GPCRs.

An important adaptive response of the cell against multiple extracellular stimuli is receptor desensitization, which refers to the reduction of receptor responsiveness despite continuing agonist stimulation. GPCRs have developed elaborate means of turning off signal (4). One mechanism for desensitization is receptor down-regulation, which refers to the net loss of receptors from the cell by a decrease in receptor synthesis, a destabilization of receptor mRNA, or an increase in receptor degradation (4, 5).

Two major patterns of down-regulation have been characterized; homologous (or agonist specific) down-regulation and heterologous (or agonist nonspecific) down-regulation (6). The homologous down-regulation indicates that stimulation of one GPCR over time by the agonist reduces expression levels of the same GPCR, without substantial effect on other GPCRs present in the same cell. In contrast, heterologous down-regulation indicates that stimulation of one GPCR reduces expression levels of different GPCR. Although the molecular mechanism of homologous down-regulation has been well analyzed using β-adrenergic receptors (βARs) (4, 5), the mechanism(s) underlying heterologous down-regulation is largely unknown.

Angiotensin (Ang) II is a major bioactive polypeptide, and improper regulation of Ang II induces various cardiovascular diseases, including hypertension, cardiac hypertrophy, fibrosis, apoptosis, and arrhythmia (7). Most of known physiological and pathological effects of Ang II are mediated via the angiotensin type 1 receptor (AT₁R). In fact, aberrant expression of AT₁R has been shown to have pathophysiological relevance in cell culture, animal studies, and clinical interventional trials (8). Ang II decreases AT₁R expression level through destabilization of AT₁R mRNA (9), suggesting the involvement of homologous down-regulation processes. In addition, the expression of AT₁Rs could be regulated by various factors, including cytokines, growth factors, and reactive oxygen species (ROS). However, heterologous down-regulation of AT₁Rs through different GPCRs has not been reported.

It is well established that nitric oxide (NO) regulates a diverse array of signal transduction pathways, acting in significant part through the covalent modification of cysteine (Cys) thiols (S-nitrosylation) that are found at active or allosteric sites of proteins (10). Several transmembrane-spanning receptors and their downstream signaling molecules could be modified by NO, including NMDA receptor (11), GPCR kinase 2 (12), β-arrestin (13), and G proteins (14, 15). Although exposure of excessive NO has been reported to decrease AT₁R density (16), the molecular details and its physiological significance are still unknown.

During the analysis of the roles of Ca²⁺ signaling in cardiac fibroblasts, we noticed that the pretreatment with adenosine triphosphate (ATP) selectively suppressed AT₁R-stimulated Ca²⁺ response. We have found that stimulation of P2Y₂ receptor with ATP down-regulates AT₁R signaling through expression of inducible NO synthase (iNOS) without nuclear factor κB (NF-κB) activation. Down-regulation of AT₁R requires functional interaction of NF-κB with iNOS, and the locally generated NO mediates cysteine modification (S-nitrosylation) of NF-κB, leading to suppression of AT₁R transcription rate.

Results

P2Y₂ Receptor Stimulation Selectively Down-Regulates AT₁R Signaling.

In a normal heart, two-thirds of the cell population is composed of cardiac fibroblasts. Expression level of AT₁R in rat neonatal cardiac fibroblasts is more than fivefold higher than that in rat neonatal cardiomyocytes (65 ± 12 fmol/mg protein), and AT₁R signaling in cardiac fibroblasts has been implicated in the development of cardiac fibrosis (17). We have previously reported that treatment with Ang II increases activity of nuclear factor of activated T cells (NFAT), a Ca²⁺-dependent transcriptional factor that is predominantly regulated by calcineurin, both in cardiac myocytes and fibroblasts (18, 19). Activation of NFAT has been implicated in the development of cardiac hypertrophy (20), but the role of NFAT in cardiac fibroblasts is not fully known. Therefore, we first examined the relationship between NFAT activity and AT₁R signaling in cardiac fibroblasts. We found that treatment with ATP more potently increased NFAT activity than Ang II, the endogenous ligand of G_qPCR (Fig. 1 A and B). Purinergic receptors are classified into two families, P2X and P2Y (21). P2X receptors are ligand-gated channels. P2Y receptors are G protein-coupled receptors, and divided into eight subtypes. Purinergic signaling has been implicated in inflammatory responses of various systems (22, 23), and we have also reported that extracellular nucleotides trigger pressure overload-induced cardiac fibrosis in mice (24). The ATP-induced increases in intracellular Ca²⁺ concentration ([Ca²⁺]_i) and NFAT activity were completely suppressed by the inhibition of phospholipase C (PLC) and knockdown of P2Y₂ receptor (P2Y₂R), but not by inhibition of P2Y₁R and P2Y₆R (Fig. S1). These results suggest that P2Y₂R predominantly regulates ATP-induced activation of Ca²⁺ signaling in rat cardiac fibroblasts.

Fig. 1.

ATP decreases AT₁R density through NFAT activation. (A) Concentration-dependent NFAT activation induced by ATP and Ang II in rat neonatal cardiac fibroblasts. Cells were treated with indicated concentration of ATP or Ang II for 6 h. (B) Time courses of NFAT activation induced by ATP and Ang II. (C) Changes in AT_1AR mRNA expression levels by the treatment with ATP (100 μM) for 24 h or overexpression of NFAT-CA in the presence or absence of CysA (100 ng/mL). Cells were treated with CysA 10 min before ATP stimulation. (D) Changes in AT₁R density induced by ATP (100 μM) for 24 h with or without CysA (n = 4–6). (E) Peak increases in [Ca²⁺]_i induced by Ang II stimulation in ATP-treated or ATP/CysA-treated cardiac fibroblasts. Cells were treated with CysA (100 ng/mL) 10 min before stimulation of ATP (100 μM) for 24 h. (F) Peak [Ca²⁺]_i increases induced by Ang II in LacZ (control)- and NFAT-CA-expressing cardiac fibroblasts. Cells were infected with Ad-LacZ or Ad-NFAT-CA for 48 h (n = 63–99). **P < 0.01.

The treatment with ATP for 24 h resulted in a decrease in expression levels of AT₁R mRNA and protein (Fig. 1 C and D), which was abolished by cyclosporine A, a specific calcineurin inhibitor. In addition, expression of Cain, a specific inhibitory peptide of calcineurin (25), also canceled the ATP-induced decrease in AT₁R density and AT₁R-induced signaling (Fig. S2 A–C). Expression of constitutively active mutant of NFAT4 (NFAT-CA) inhibited AT₁R mRNA expression (Fig. 1C), indicating that calcineurin-NFAT signaling mediates AT₁R down-regulation. As AT₁R couples with G_q family proteins, the Ang II-induced transient increase in [Ca²⁺]_i was measured as an index of the magnitude of AT₁R signaling. Pretreatment with ATP or overexpression of NFAT-CA significantly decreased the Ang II-induced maximal [Ca²⁺]_i increases without changing EC₅₀ values (Fig. 1 E and F). There are four NFAT isoforms responsive to increase in [Ca²⁺]_i (26). Among them, NFATc1 participates in cardiac valve development and NFATc3 participates in the development of cardiac hypertrophy. Treatment of cardiac fibroblast with ATP induced nuclear translocation of NFATc1 and NFATc3, but not NFATc2 (Fig. S2 D and E). We could not detect the expression of NFATc4 proteins. These results suggest that stimulation of G_q-coupled P2Y₂ receptor by ATP induces heterologous down-regulation of G_q-coupled AT₁R through activation of NFATc1 and NFATc3 in rat cardiac fibroblasts. Although P2Y₂R can be stimulated not only by ATP but also by ADP and UTP (21), ATP maximally increased NFAT activity and decreased AT₁R signaling (Fig. S3). However, endothelin-1 and bradykinin, which signal through G_q (27), failed to imitate ATP both in NFAT activation and inhibition of G_q signaling, suggesting that the ATP-mediated suppression of AT₁R signaling is likely to correlate with its potential to increase NFAT activity. Furthermore, pretreatment with Ang II did not affect the ATP-induced Ca²⁺ response (Fig.S3 C and D), suggesting that ATP specifically induces heterologous down-regulation of AT₁R signaling but not vice versa.

ATP Decreases AT₁R Density Through iNOS Expression.

We next examined the mechanism of AT₁R down-regulation by NFAT activation. Because the mRNA stability of AT₁R was not decreased by ATP treatment (Fig. S3E), its effects on the AT₁R transcription rate were further studied. NF-κB has been reported to participate in AT₁R up-regulation induced by cytokines or bacterial toxin (28, 29), and inhibition of NF-κB significantly suppressed AT₁R transcription rate and AT₁R density (Fig. S4). Thus, basal activity of AT₁R transcription may be mainly regulated by NF-κB. In addition, NO has been reported to decrease NF-κB and AT₁R transcription (16). Thus, NO may mediate ATP-induced AT₁R down-regulation. Treatment with ATP induced a time-dependent decrease in expression of luciferase from AT₁R promoter, as well as a decrease in NF-κB–dependent luciferase activity (Fig. 2A). The ATP-induced suppression of AT₁R transcription rate was abolished by N(G)-nitro-l-arginine methyl ester (l-NAME) or N-(3-(aminomethyl)benzyl)acetamidine (1400W), a specific inhibitor of iNOS (Fig. 2B). Treatment with 1400W canceled the ATP-induced decrease in AT₁R density and suppression of AT₁R-stimulated Ca²⁺ response (Fig. 2 C–E). In addition, knockdown of iNOS, but not endothelial NOS (eNOS), abolished the ATP-induced suppression of Ca²⁺ response by AT₁R stimulation (Fig. 2F), suggesting that iNOS mediates ATP-induced AT₁R down-regulation. It is generally accepted that NF-κB predominantly regulates the expression level of iNOS proteins (30). However, it has also been reported that iNOS proteins are induced by the constitutive activation of calcineurin-NFAT signaling in cardiomyocytes (31). In accordance with this, the expression of a constitutively active mutant of NFAT (NFAT-CA) decreased AT₁R-stimulated Ca²⁺ response, which was completely abolished by the treatment with l-NAME or 1400W (Fig. S5). ATP-stimulated increase of iNOS proteins was completely suppressed by 1400W (Fig. 2 G and H). In addition, ATP-induced NO production was completely suppressed by the knockdown of iNOS, but not eNOS (Fig. 2I). As ATP strongly induced NFAT activity, NFAT-dependent iNOS expression may mediate the ATP-induced down-regulation of AT₁R in cardiac fibroblasts.

Fig. 2.

iNOS mediates ATP-induced AT₁R down-regulation. (A) Time-dependent suppression of NF-κB-dependent and rAT_1AR promoter-dependent luciferase activities induced by ATP (100 μM). (B) Effects of iNOS inhibitors on the ATP-induced decrease in rAT_1AR promoter activity. Cells were treated with l-NAME (100 μM) or 1400W (10 μM) 10 min before sitmulation with ATP (100 μM) for 24 h. (C) Effects of 1400W on the ATP-induced decrease in AT₁R protein expression (n = 3–5). (D) Time courses of AT₁R-stimulated increases in [Ca²⁺]_i in cardiac fibroblasts pretreated with ATP or ATP and 1400W. (E) Peak increases in [Ca²⁺]_i induced by Ang II (100 nM). Cells were treated with 1400W (10 μM) 10 min before sitmulation with ATP (100 μM) for 24 h. (F) Peak increases in [Ca²⁺]_i induced by Ang II in siRNA-transfected cardiac fibroblasts pretreated with or without ATP (100 μM) for 24 h. Cells were treated with ATP 48 h after transfection with siRNAs for eNOS and iNOS. (n = 42–78) (G and H) Changes in expressions of iNOS, eNOS and GAPDH proteins by the stimulation with ATP and IL-1β (1 ng/mL) for 24 h in the presence or absence of 1400W (10 μM). (I) Effects of siRNAs for eNOS or iNOS on the ATP-induced nitrite production. Cells were treated with ATP (100 μM) for 24 h. (n = 3). *P < 0.05, **P < 0.01.

Requirement of S-Nitrosylation of p65 for ATP-Induced AT₁R Down-Regulation.

As the treatment with protein kinase G (PKG) inhibitor (KT5823) did not cancel the ATP-induced down-regulation of AT₁R signaling, and a cGMP analog (8-bromo-cGMP) did not induce AT₁R down-regulation (Fig. S5 A and B), NO may suppress AT₁R transcription in a cGMP-independent manner. S-nitrosylation has been recently recognized as a new NO-based but cGMP-independent signaling regulating several cardiovascular functions (10). Dimers of NF-κB proteins are composed of five transcriptional factors: p50, p52, p65, Rel, and RelB, which all share an N-terminal Rel homology domain (32). Among them, S-nitrosylation of p50 and p65 has been reported to participate in iNOS-induced decrease in NF-κB transcriptional activity (33, 34). We next examined whether S-nitrosylation participates in ATP-induced AT₁R down-regulation. It was reported that a conserved Cys within the DNA-binding site of the Rel homology domain is the site of S-nitrosylation (33). Thus, we constructed Cys mutants of p65 (p65-C38S) and p50 (p50-C62S) to examine which Cys is involved in ATP-induced AT₁R down-regulation. Overexpression of wild-type p65 (p65-WT) and p65-C38S enhanced the AT₁R-stimulated increases in [Ca²⁺]_i, suggesting that transcriptional activity of p65 is not affected by substitution of Ser for Cys. The ATP-induced suppression of AT₁R signaling normally occurred in vector- and p65-WT–expressing cardiac fibroblasts but was completely abolished in p65-C38S–expressing cardiac fibroblasts (Fig. 3 A and B). Substitution of Ser for other eight Cys residues of p65 did not affect the ATP-induced suppression of AT₁R signaling (Fig. S6), suggesting that Cys38 of p65 participates in ATP-induced S-nitrosylation and AT₁R down-regulation. In contrast, expression of p50-C62S did not abolish the suppression of AT₁R signaling induced by ATP (Fig. 3 C and D). Although p50 essentially lacks transactivation domain (32), overexpression of p50-WT and p50-C62S enhanced the AT₁R-stimulated Ca²⁺ responses, suggesting that p50 does not mediate ATP-induced suppression of AT₁R signaling. It has been previously reported that phosphorylation of p65 at Ser468 reduces basal NF-κB activity (35). However, treatment with ATP did not increase the phosphorylation level of p65 at Ser468, and substitution of Ala for Ser468 did not affect the suppression of AT₁R-stimulated Ca²⁺ response induced by ATP (Fig. S6 D–F). Thus, phosphorylation of p65 is not involved in ATP-induce AT₁R down-regulation. The expression of iNOS proteins induced S-nitrosylation of p65 at Cys38 in HEK293 cells (Fig. 3E), and the treatment of cardiac fibroblasts with ATP induced S-nitrosylation of p65 in an iNOS-dependent manner (Fig. 3F). These results suggest that S-nitrosylation of p65 at Cys38 participates in the ATP-induced AT₁R down-regulation in cardiac fibroblasts.

Fig. 3.

S-nitrosylation of p65 mediates ATP-induced AT₁R down-regulation. (A) Time courses of Ca²⁺ responses induced by Ang II (100 nM) in vector, p65 (WT) or p65 (C38S)-overexpressing cardiac fibroblasts. Cells were pretreated with ATP (100 μM) for 24 h. (B) Peak increases in [Ca²⁺]_i induced by Ang II (n = 41–68). (C, D) Time courses (C) and peak changes (D) of Ca²⁺ responses induced by Ang II in vector, p50 (WT) or p50 (C62S)-expressing fibroblasts (n = 44–72). (E) S-nitrosylation of Cys38 of p65 in p65 and iNOS-expressing HEK293 cells. Cell lysates were incubated with ascorbate (1 mM) for 1 h at 25 °C in the dark room. (F) S-nitrosylation of p65 and iNOS expression induced by ATP stimulation in cardiac fibroblasts. Cells were treated with 1400 W (10 μM) 10 min before stimulation with ATP (100 μM) for 24 h. Cell lysates were incubated with ascorbate (1 mM) for 1 h at 25 °C in the dark room (n = 3). *P < 0.05, **P < 0.01.

Requirement of the Interaction Between iNOS and p65 for ATP-Induced AT₁R Down-Regulation.

The extent of iNOS induction by ATP was smaller than that by IL-1β, a potent activator of NF-κB signaling (Fig. 2 G and H). This apparent discrepancy may be explained by the differences in the colocalization between p65 and iNOS, which is predominantly expressed in the cytosol (33), in ATP-, and IL-1β–treated cells. In iNOS- and p65-overexpressing HEK293 cells, both p65-WT and p65-C38S interacted with iNOS (Fig. 4A). In addition, iNOS induced by ATP, but not eNOS, interacted with p65 in rat cardiac fibroblasts, which was completely suppressed by cyclosporine A (Fig. 4B). Although cotreatment with S-nitrosoglutathione (GSNO) (1 mM) significantly enhanced ATP-induced S-nitrosylation of p65, GSNO did not affect the ATP-induced interaction between p65 and iNOS (Fig. S7). These results suggest that p65 associates with iNOS independently of Cys modification. In the resting condition, p65 was expressed both in the cytosol and nucleus, whereas iNOS was hardly detected in cardiac fibroblasts (Fig. 4C). Treatment with ATP induced expression of iNOS in the cytosol without affecting the subcellular localization of p65, resulting in colocalization of p65 and iNOS in the cytosol. In contrast, IL-1β markedly increased expression levels of iNOS and induced continuous translocation of p65 into the nucleus, resulting in little colocalization between p65 and iNOS. We next examined which domain of iNOS is required for the interaction with p65. As iNOS has N-terminal oxygenase domain and C-terminal reductase domain (36), and the C-terminal domain has flavin-binding domain and NADPH-binding domain, we constructed four fragments of iNOS (Fig. 4 D and E). In HEK293 cells expressing p65 and each iNOS fragment (Fr-I–Fr-IV), p65 was interacted with Fr-III only, which contains flavin-binding domain (Fig. 4F). Furthermore, expression of cardiac fibroblasts with Fr-III completely blocked the ATP-induced interaction of p65 with iNOS and suppression of AT₁R-stimulated Ca²⁺ response (Fig. 4 G and H). These results suggest that iNOS induced by ATP stimulation interacts with p65 in the cytosol through flavin-binding domain, which is required for S-nitrosylation of p65 and which decreases its activity in rat cardiac fibroblasts.

Fig. 4.

ATP-induced iNOS interacts with p65. (A) Association of iNOS proteins with p65 (WT or C38S) in p65 and iNOS-expressing HEK293 cells. IP: immunoprecipitation, WB: Western blotting. HA-tagged iNOS proteins were immunoprecipitated with anti-HA antibody (IP), and combined p65 proteins were detected by Western blotting. Expression levels of p65 and iNOS were also confirmed using total cell lysates (lysates). (B) ATP stimulation-dependent interaction of p65 with iNOS in cardiac fibroblasts. Cells were treated with ATP (100 μM) for 24 h in the presence or absence of CysA (100 ng/mL). Native p65 proteins were immunoprecipitated with anti-p65 antibody (IP) (n = 3). (C) Localization of p65 and iNOS proteins in cardiac fibroblasts stimulated with ATP (100 μM) or IL-1β (1 ng/mL) for 24 h. (D) Structure of iNOS fragments (Fr-I – Fr-IV). (E) Expression of flag-tagged iNOS fragments. FL; full length. (F) Interaction of p65 with Fr-III. (G) Effects of Fr-III on the ATP-induced interaction of p65 with iNOS (n = 3). (H) Effects of iNOS fragments on the ATP-induced down-regulation of AT₁R signaling (n = 39–62). *P < 0.05.

Previous reports have clearly shown that β-arrestins associate with NOS and IκBα (13, 37–39). As the IκBα interacts with NF-κB p65 subunit, ATP may induce the formation of quaternary complex of iNOS/β-arrestin/IκBα/p65 in cardiac fibroblasts. To verify this hypothesis, we further examined whether knockdown of β-arrestin(s) cancels the ATP-induced p65-iNOS interaction. Knockdown of β-arrestin2 significantly suppressed the down-regulation of AT₁R signaling (Fig. S8). The β-arrestins were predominantly expressed in the cytosol of cardiac fibroblasts, and the ATP-induced colocalization of p65 with iNOS was diminished by β-arrestin2 knockdown. In addition, p65 actually interacted not only with iNOS, but also IκBα and β-arrestin2 in ATP-pretreated cardiac fibroblasts, which was diminished by β-arrestin2 knockdown. Although knockdown of β-arrestin1 also suppressed the p65-iNOS colocalization and down-regulation of AT₁R signaling induced by ATP pretreatment, the magnitude of suppression was smaller than that of β-arrestin2. These results indicate that β-arrestins mediate the ATP-induced formation of quaternary complex of iNOS/β-arrestins/IκBα/p65 in the cytosol of cardiac fibroblasts. Although β-arrestins participate in GPCR-stimulated activation of extracellular signal-regulated kinase (ERK), treatment with ERK inhibitors (U0126 and PD98059) did not cancel the ATP-induced down-regulation of AT₁R signaling (Fig. S9). Thus, β-arrestins may participate in ATP-induced AT₁R down-regulation through anchoring the interaction between p65 and iNOS proteins.

ATP Mediates Down-Regulation of AT₁R Signaling in Pressure-Overloaded Mouse Hearts.

We finally examined whether ATP-induced AT₁R down-regulation occurred in vivo. We have previously reported that nucleotides (such as ATP and UDP) released from cardiac myocytes stimulate the production of fibrotic genes, which activate the production of collagen in cardiac fibroblasts in pressure overload-induced mice (24). As nucleotides are released from cardiac myocytes and the expression levels of AT₁Rs in cardiac fibroblasts are higher than those in cardiac myocytes, both cardiac myocytes and fibroblasts were cocultured on a silicone rubber dish. Mechanical stretch that stimulates nucleotides release from cardiac myocytes (24) decreased AT₁R density, which was canceled by pretreatment with l-NAME (Fig. 5A). As expected, inhibition of P2Y₂R by PPADS or P2Y₂R siRNA completely abolished mechanical stretch-induced AT₁R down-regulation. Mechanical stretch of cardiac cells induced ATP release, which leads to NO production in P2Y₂R and iNOS-dependent manners (Fig. 5 B and C). Furthermore, 6 wk of transverse aortic constriction (TAC) induced significant increases in iNOS mRNA and protein levels and S-nitrosylation of p65, which were completely suppressed by the treatment with suramin, a P2 receptor antagonist (Fig. 5 D and E). Pressure overload also increased expression levels of eNOS proteins, but this increase was not affected by suramin. In addition, 6 wk of TAC decreased expression levels of AT₁R mRNA and proteins, which were also abolished by suramin treatment (Fig. 5 E and F). Furthermore, pressure overload-induced S-nitrosylation of p65 was diminished in iNOS knockout mouse hearts (Fig. 5 G and H) and in 1400W-treated mouse hearts (Fig. S10). These results suggest that mechanical stretch-induced ATP release causes AT₁R down-regulation through iNOS-dependent S-nitrosylation of p65 in pressure overloaded mice.

Fig. 5.

Mechanical stretch induces AT₁R down-regulation through ATP-induced S-nitrosylation of NF-κB. (A) Effects of l-NAME (100 μM), d-NAME (100 μM), PPADS (100 μM), and P2Y₂R siRNA on mechanical stretch-induced decrease in AT₁R density in rat cardiac myocytes. (B) Effects of 1400W (10 μM) and P2Y₂R siRNA on mechanical stretch-induced nitrite production. (C) Effects of 1400W on mechanical stretch-induced ATP release (n = 3–5). (D–F) S-nitrosylation of p65, changes in expression of iNOS, p65, eNOS and AT₁R proteins induced by pressure overload in the presence or absence of suramin in mouse hearts. (E) Changes in iNOS and AT_1AR mRNAs induced by TAC. (F) Changes in AT₁R density induced by TAC in the presence or absence of suramin (n = 3–6). (G and H) S-nitrosylation of p65 induced by pressure overload in wild type (+/+) and iNOS-knockout (−/−) mouse hearts. Tissue lysates were treated with or without ascorbate (1 mM) for 1 h (n = 3). *P < 0.05, **P < 0.01.

Discussion

A large number of studies have shown that a wide range of physiological and pathological stimuli modulate AT₁R expression in a number of cell types and tissues (40). We first demonstrated that ATP induces down-regulation of AT₁R signaling in rat cardiac fibroblasts. Extracellular ATP in the cardiovascular system may be originated from various cellular sources: perivascular sympathetic nerve endings, myocytes, endothelial cells, and inflammatory cells (14). We found that mechanical stretch-induced ATP release decreases AT₁R density in mouse hearts with pressure overload (Fig. 5). Although both ATP and Ang II are believed to function as an inflammatory factor (41), our results suggest that P2Y₂R-stimulated iNOS expression negatively regulates Ang II-mediated inflammatory response of the heart. We have previously reported that nucleotides released from cardiac myocytes activate P2Y₆R of cardiac myocytes and induce fibrotic genes (24). In addition, Braun et al. has recently reported that P2Y₂R stimulation by UTP induces fibrotic responses of cardiac fibroblasts (42). Thus, extracellular nucleotides contribute to cardiac fibrosis at least two independent pathways: one is induction of fibrotic genes in cardiac myocytes through P2Y₆R and another is iNOS-mediated signaling in cardiac fibroblasts through P2Y₂R. The pathophysiological roles of P2Y₂R signaling in the heart are still unclear, but our findings will provide a unique insight into the cross-talk between purinergic signaling and Ang II signaling in cardiovascular systems.

We revealed that local production of NO induces S-nitrosylation of NF-κB p65 and suppresses AT₁R transcription leading to down-regulation of AT₁R. Several reports have sugested the involvement of S-nitrosylation in NO-dependent (but cGMP independent) regulation of β-adrenergic receptor desensitization (12, 13). Most studies have shown the involvement of S-nitrosylation in GPCR desensitization by exposing excessive concentrations of NO to the cell. However, GRK2 is endogenously S-nitrosylated and levels are lower in eNOS null animals and higher in GSNOR^−/− animals. Furthermore, receptor internalization in cellular assays and desensitization in situ are regulated by endogenously derived NO. We also found that endogenous S-nitrosylation of NF-κB p65 requires the formation of iNOS/β-arrestins/IκBα/p65 quaternary complex in the cytosol. This finding emphasizes the fact that β-arrestins play a key mediator in NO-based signaling. Although IL-1β potently induces expression of iNOS, IL-1β did not suppress NF-κB activity. This may be explained by the evidence that p65 was not colocalized with iNOS by IL-1β stimulation (Fig. 4). Thus, the complex formation between NO donor and acceptor is a critical factor of spatiotemporal regulation of NO signaling induced by ATP stimulation.

In conclusion, we reveal that S-nitrosylation of p65 is required for heterologous down-regulation of G_q-coupled AT₁R by another G_q-coupled P2Y₂R stimulation. Although it is generally thought that heterologous regulation of GPCRs is mediated by second messenger-activated kinases, heterologous down-regulation of AT₁R by ATP did not require the kinase activation in which cysteine modification of transcriptional factor plays essential roles. Our findings will provide a unique insight into cross-talk between GPCR signaling pathways.

Materials and Methods

Materials, recombinant adenoviruses, and culture of cardiac fibroblasts, measurement of NF-κB-luciferase and AT₁R-luciferase activity, measurement of extracellular ATP concentration and NO production, measurement of AT₁R and iNOS expressions, quantification of intracellular Ca²⁺ concentration, S-nitrosylation biotin switch assay, immunoprecipitation, and confocal visualization of NF-κB p65 subunit and iNOS proteins are described in SI Materials and Methods.

Animals and TAC Surgery.

All protocols using mice and rats were approved by the guidelines of Kyushu University. Mice with a homozygous deletion of the iNOS gene were purchased from The Jackson Laboratory. TAC surgery was performed on 6-wk-old male C57BL/6J mice (24). A mini-osmotic pump (Alzet) filled with vehicle (saline), suramin or 1400W was implanted intraperitoneally 3 d after TAC into 6-wk-old male C57BL/6J mice.

Statistical Analysis.

The results are presented as mean ± SEM from at least three independent experiments. The representative data of time course experiments were plotted from one of three similar experiments that were performed with more than 20 cells. Statistical comparisons were made with two-tailed Student's t test or one way analysis of variance followed by Student-Newman-Keuls procedure, with significance imparted at P < 0.05.