Distinct Roles of Protease-Activated Receptors in Signal Transduction Regulation of Endothelial Nitric Oxide Synthase

Abstract

Protease-activated receptors such as PAR1 and PAR2 have been implicated in the regulation of endothelial nitric oxide production. We hypothesized that PAR1 and PAR2 distinctly regulate the activity of endothelial nitric oxide synthase through the selective phosphorylation of a positive regulatory site, Ser¹¹⁷⁹ and a negative regulatory site, Thr⁴⁹⁷ in bovine aortic endothelial cells. A selective PAR1 ligand, TFLLR, stimulated the phosphorylation of endothelial nitric oxide synthase at Thr⁴⁹⁷. It had a minimal effect on Ser¹¹⁷⁹ phosphorylation. In contrast, a selective PAR2 ligand, SLIGRL, stimulated the phosphorylation of Ser¹¹⁷⁹ with no noticeable effect on Thr⁴⁹⁷. Thrombin has been shown to transactivate PAR2 through PAR1. Thus, thrombin as well as a peptide mimicking the PAR1 tethered ligand, TRAP, stimulated phosphorylation of both sites. Also, thrombin and SLIGRL, but not TFLLR, stimulated cyclic GMP production. A G_q inhibitor blocked thrombin- and SLIGRL-induced Ser¹¹⁷⁹ phosphorylation whereas it enhanced thrombin-induced Thr⁴⁹⁷ phosphorylation. In contrast, a G_12/13 inhibitor blocked thrombin- and TFLLR-induced Thr⁴⁹⁷ phosphorylation whereas it enhanced the Ser¹¹⁷⁹ phosphorylation. Although a Rho-kinase inhibitor, Y27632, blocked the Thr⁴⁹⁷ phosphorylation, other inhibitors that targeted Rho-kinase failed to block TFLLR-induced Thr⁴⁹⁷ phosphorylation. These data suggest that PAR1 and PAR2 distinctly regulate endothelial nitric oxide synthase phosphorylation and activity through G_12/13 and G_q, respectively, delineating the novel signaling pathways by which the proteases act on protease-activated receptors to potentially modulate endothelial functions.

Introduction

Thrombin is a multifunctional serine protease that not only mediates the coagulation cascade but also has a wide variety of actions within the endothelium and vascular smooth muscle. The signal transduction and functions of thrombin through the protease-activated receptors (PARs) are strongly implicated in vascular physiology and pathophysiology ^{1, 2}. PARs represent a unique class of G protein-coupled receptors (GPCRs) activated by proteolytic cleavage of their extracellular N-terminal domains. The new N-terminus acts as a “tethered ligand” and activates the receptor itself. Four members of this family have been cloned, of which PAR1 and PAR3 are essentially thrombin receptors, while PAR2 is activated by trypsin or mast cell tryptase, and PAR4 is activated by both thrombin and trypsin ¹. However, transactivation of PAR2 by thrombin through PAR1 (the tethered ligand of PAR1 acts as a PAR2 ligand to activate PAR2 if both receptors co-exist) has been demonstrated in cultured human umbilical vein endothelial cells (HUVECs) ³. Activation of PARs stimulate a myriad arrays of signal transduction pathways, which include the activation of distinct heterotrimeric G proteins and tyrosine and serine/threonine kinases ^{1, 2}.

Importantly, multiple PARs seem to regulate the release of nitric oxide (NO) via endothelial NO synthesis (eNOS) ^4-6. Although these data suggest a significant link between eNOS and PARs, detailed mechanisms by which PARs regulate eNOS activity have been insufficiently characterized. Recent studies revealed that eNOS needs to be specifically phosphorylated to exert its full activity ⁷. There are multiple phosphorylation sites on eNOS, however most is known about the functional consequences of phosphorylation of Ser^{1179 (bovine)/1177 (human)} and Thr^{497 (bovine)/495 (human)}. When the Ser¹¹⁷⁹ is phosphorylated, NO production is increased two- to threefold above basal level. In contrast, Thr⁴⁹⁷ is a negative regulatory site of eNOS associated with decreased enzymatic activity. Several eNOS kinases such as Akt, which phosphorylates Ser¹¹⁷⁹ or protein kinase C (PKC), which phosphorylates Thr⁴⁹⁷ have been identified ⁷.

Here, we have hypothesized that PAR1 and PAR2 distinctly regulate the two phosphorylation sites on eNOS and NO production. We found several lines of evidence indicating the reciprocal regulation of the phosphorylation of these sites on eNOS by the PARs in bovine aortic endothelial cells (BAECs) through different G protein-dependent signal transduction pathways. These data demonstrate novel signal transduction characteristics of PAR1 and PAR2 in endothelial cells representing a diverse physiological and pathophysiological role of PARs in regulating endothelial functions.

Materials and Methods

Reagents

Thrombin from bovine plasma was purchased from Sigma. A PAR1 selective agonist, TFLLR-NH₂, a PAR4 selective agonist, AY-NH₂, and a peptide mimicking the PAR1 tethered ligand, TRAP, were purchased from Tocris. A PAR2 selective agonist, SLIGRL-NH₂, was purchased from Bachem. A G_q inhibitor, YM-254890, was a gift from Astellas Pharma Inc. Pertusis toxin, phorbol 12-myristrate 13-acetate (PMA), and Rho-kinase (ROCK) inhibitors, Y27632 and H-1152, and a PKC inhibitor, GF109203X, were purchased from Calbiochem. Antibodies against Ser¹¹⁷⁹-phosphorylated eNOS and Thr⁴⁹⁷-phosphorylated eNOS were purchased from Cell Signaling Technology. Total eNOS antibody was purchased from BD Transduction Laboratories. Antibody for Thr⁸⁵³-phosphorylated myosin phosphatase target subunit-1 (MYPT1) was purchased from Upstate and antibody against total MYPT1 was purchased from Covance.

Cell Culture

BAECs were purchased from Cambrex and grown in DMEM containing 10% fetal bovine serum, penicillin and streptomycin as previously described ⁸. BAECs were subcultured using Versene (0.53 mmol/L EDTA in phosphate-buffered saline) to avoid trypsin exposure. HUVECs were a gift from Dr. Yi Wu (Temple University School of Medicine) ⁹. Cells from passage 4-12 were grown to about 90% confluency and serum-starved for 48 h before the experiments.

Adenovirus Infection

Generation and characterization of replication-deficient adenoviruses encoding a dominant-negative mutant of Rho (dnRho), myc-N¹⁹-RhoA, and myc-p115RGS were described in detail elsewhere ^{10, 11}. Adenovirus construct encoding GRK2/βARK-ct was generated by Dr. Andrea Eckhart from Gene Transfer Vector Core (Thomas Jefferson University). An adenovirus vector encoding GFP was used as a control for other adenovirus vectors which also encode GFP independently of their respective inserts. The adenovirus titers were determined by Adeno-X™ Rapid Titer Kit (BD Biosciences). Subconfluent BAECs were infected with adenovirus for 2 days as previously described ⁸.

Immunoblotting

Cell lysates were subjected to SDS-PAGE gel electrophoresis and electrophoretically transferred to a nitrocellulose membrane as previously described ¹². The membranes were then exposed to primary antibodies overnight at 4°C. After incubation with the peroxidase linked secondary antibody for 1 h at room temperature, immunoreactive proteins were visualized by ECL reagent (Pierce).

Intracellular cyclic GMP Assay

BAECs were incubated with agonists at 37°C for 20 min in the presence of 0.5 mmol/L methylisobutylxanthine and intracellular cyclic GMP (cGMP) was determined by an enzyme immunoassay kit (Cayman Chemical) ⁸.

Intracellular Ca²⁺ ([Ca²⁺]_i) Measurements

[Ca²⁺]_i was measured as described previously using fura 2 as an indicator ¹³. In brief, BAECs subcultured on coverslips were loaded with 3 μM fura 2-AM. The fura 2 fluorescence was measured at a frequency of 1 Hz and [Ca²⁺]_i-values were then obtained as described ¹³.

Statistical Analysis

The data shown are either from 3 or 4 independent experiments (the total n number is given in the figure legends) and presented as mean±SEM. The data were analyzed using one-way ANOVA followed by a post-hoc modified t-test or a two-way ANOVA. P<0.05 was considered significant.

Results

eNOS Phosphorylation by Thrombin in BAECs

In cultured BAECs, we have examined whether thrombin stimulates phosphorylation of eNOS at Ser¹¹⁷⁹, a catalytically positive regulatory site, and at Thr⁴⁹⁷, a catalytically negative regulatory site ⁷. As shown in Figure 1A, thrombin (10 U/mL) stimulated eNOS phosphorylation at Ser¹¹⁷⁹, which began at 0.5 min, peaked at 1 min, and phosphorylation declined thereafter. Also, thrombin stimulated eNOS phosphorylation at Thr⁴⁹⁷ from 0.5 to 2 min. Figure 1B shows the concentration dependence of thrombin-induced phosphorylation of eNOS in BAECs. eNOS phosphorylation at Ser¹¹⁷⁹ was detectable with 1U/mL of thrombin, and the maximum phosphorylation was observed at a concentration of 10 U/mL. In contrast, thrombin-stimulated phosphorylation at Thr⁴⁹⁷ occurred at an even lower concentration of thrombin (0.1 U/mL) in BAECs. Thus, these results suggest the possibility that thrombin mediates the phosphorylation of each eNOS site through distinct PARs or G proteins in BAECs.

Figure 1.

The effects of thrombin stimulation on eNOS phosphorylation at Ser¹¹⁷⁹ and Thr⁴⁹⁷ were investigated in BAECs. The cells were stimulated with 10 U/mL of thrombin for the indicated periods (A) or with indicated concentrations of thrombin for 1 min (B), and phosphorylation of eNOS at each site was determined using immunoblot analysis. The bar graphs show densitometric analyses of the phosphorylation of eNOS at each site (mean ± SEM, *p < 0.05 compared with the basal control, n=3).

eNOS Phosphorylation and Activation by Selective PAR Agonists in BAECs

To examine the participation of PARs in the regulation of NO production, we stimulated BAECs with a selective PAR1 agonist, TFLLR, or a selective PAR2 agonist, SLIGRL, and evaluated eNOS phosphorylation. As shown in Figure 2A, TFLLR stimulated eNOS phosphorylation at Thr⁴⁹⁷ at 1 to 2 min, whereas it had no effect on the Ser¹¹⁷⁹ phosphorylation. On the other hand, SLIGRL stimulated eNOS phosphorylation at Ser¹¹⁷⁹ but not Thr⁴⁹⁷ at 1 to 2 min. In line with a theory that thrombin activates PAR2 via the intermolecular transactivation of PAR2 ³, a peptide mimicking the PAR1 tethered ligand, TRAP, stimulated both phosphorylation sites in a concentration dependent manner (10-100 μmol/L). However, a PAR4 agonist, AY-NH₂ (up to 200 μmol/L), had no effect on either phosphorylation site (data not shown). The distinct responses by the PAR1 and PAR2 agonists were also confirmed in HUVECs and bovine pulmonary artery endothelial cells (data not shown). These results suggest that PAR1 and PAR2 have distinct roles in regulating eNOS phosphorylation at least in these types of endothelial cells.

Figure 2.

Distinct phosphorylation responses of eNOS stimulated by PAR1 and PAR2. A. The effects of a PAR1 or PAR2 agonist on eNOS phosphorylation. BAECs were stimulated with a PAR1 agonist, TFLLR (50 μmol/L), a PAR2 agonist, SLIGRL (50 μmol/L) for 2 min, or thrombin (10 U/mL) for the indicated periods. The bar graphs show densitometric analyses of the phosphorylation of eNOS at each site (mean ± SEM, bands scanned are n=3, each, *p < 0.05 compared with the basal control). Results were expressed as percentage increase in which the maximum response to thrombin (10 U/mL) was defined as 100%, because the basal signals were more varied depending on film exposure than the stimulated signals. In B and C, the cells were stimulated with thrombin (10 U/mL), a PAR1 agonist, TFLLR (B), or a PAR2 agonist, SLIGRL (C) for 20 min at the indicated concentrations, and intracellular cGMP production was determined. The data shown are mean ± SEM, (*p < 0.05 compared with the basal control, cGMP values shown are n=4, each).

To assess eNOS activation by the PAR agonists, intracellular cGMP production was measured as a marker of NO production after stimulation of BAECs by thrombin or PAR agonists. Although thrombin stimulated intracellular cGMP production (basal 3.55±0.36 pmol/well vs. thrombin 17.50±2.53 pmol/well), a PAR1 agonist, TFLLR, at concentrations from 50 to 200 μmol/L did not stimulate cGMP production (Figure 2B). In contrast, a PAR2 agonist, SLIGRL, stimulated intracellular cGMP production in a concentration-dependent manner from 2 to 50 μmol/L (Figure 2C). Also, AY-NH2, a PAR4 agonist, did not increase in cGMP production (data not shown). We confirmed that PAR2 stimulated cGMP through NO production because L-NAME treatment completely blocked PAR2-induced cGMP production (data not shown). We have previously shown that L-NAME inhibited cGMP production by thrombin ⁹. These data indicate a preferential role of PAR2 in eNOS activation in BAECs.

PAR Activation of G_q is Required for Phosphorylation of eNOS at Ser¹¹⁷⁹ but not Thr⁴⁹⁷

PARs are known to couple to multiple G proteins including G_q, G_i and G_12/13 ¹. Thus, the difference in eNOS phosphorylation profile by PARs may be due to distinct G proteins coupling to PAR1 and PAR2 in endothelial cells. We have confirmed our recent observation that a selective G_q inhibitor, YM-254890 ¹⁴, markedly inhibited thrombin-induced eNOS Ser¹¹⁷⁹ phosphorylation ⁹, whereas it enhanced the thrombin-induced phosphorylation of eNOS at Thr⁴⁹⁷ (Figure 3A). The maximum inhibition was observed with 100-1000 nmol/L (Figure 3A and Figure S1A, please see http://hyper.ahajournals.org.). However, neither pertusis toxin (100 ng/mL, 24 h), a G_i inhibitor, nor infection of adenovirus (100 moi, 48 h) encoding GRK2-ct, a G_βγ inhibitor, blocked thrombin-induced eNOS phosphorylation at Ser¹¹⁷⁹ or Thr⁴⁹⁷ (data not shown). YM-254890 also markedly inhibited PAR2 (SLIGRL)-induced eNOS Ser¹¹⁷⁹ phosphorylation, whereas it had no inhibitory effect on PAR1 (TFLLR)-induced eNOS Thr⁴⁹⁷ phosphorylation (Figure 3B and 3C). YM-254890 also inhibited PAR2-induced cGMP production (Figure S1B) as it did in BAECs stimulated with thrombin ⁹.

Figure 3.

G_q is required for eNOS Ser¹¹⁷⁹ phosphorylation stimulated by PAR2 in BAECs. In A-C, the cells were pretreated with or without 1 μmol/L YM-254890 for 10 min, and stimulated with or without thrombin (10 U/mL) for 1 min (A), a PAR1 agonist, TFLLR (50 μmol/L) for 2 min (B), or a PAR2 agonist, SLIGRL (50 μmol/L) for 2 min (C). eNOS phosphorylation was evaluated by immunoblot analysis. The bar graphs show densitometric analyses of the phosphorylation of eNOS at each site (mean ± SEM, *p < 0.05 compared with the stimulated control, bands scanned are n=3, each).

G_q activation by PARs leads to phospholipase C-dependent rapid and transient intracellular Ca²⁺ elevation ¹. We have also demonstrated that thrombin-induced eNOS Ser¹¹⁷⁹ phosphorylation requires intracellular Ca²⁺ elevation ⁹. To test whether G_q coupling of PARs is distinct in BAECs, the effects of the PAR agonists on intracellular Ca²⁺ concentrations were examined. Both TFLLR and SLIGRL significantly elevated intracellular Ca²⁺ concentrations, which were completely inhibited by 1 μmol/L YM-254890 (data not shown). These data suggest that G_q coupling and subsequent intracellular Ca²⁺ elevation are required but not sufficient enough for eNOS Ser¹¹⁷⁹ phosphorylation as well as its activation.

G_12/13 and a Y27632 Sensitive Kinase but not ROCK are Involved in eNOS Phosphorylation at Thr⁴⁹⁷ by PAR1

PAR1 has been shown to couple to G_12/13 leading to its downstream Rho and ROCK activation in endothelial cells ¹⁵. As shown in Figure 4, a specific G_12/13 inhibitor, p115RGS, markedly inhibited thrombin- or TFLLR (PAR1)-induced eNOS Thr⁴⁹⁷ phosphorylation. It also inhibited thrombin- or TFLLR-induced ROCK activation as judged by phosphorylation of a ROCK substrate, MYPT1, at Thr⁸⁵³. Inhibition of G_12/13 leads to stimulation of Ser¹¹⁷⁹ phosphorylation by TFLLR or thrombin. p115RGS by itself had no obvious effect on either site of eNOS phosphorylation (Figure S2A). A ROCK inhibitor, Y27632, markedly inhibited thrombin- or PAR1 (TFLLR)-induced eNOS Thr⁴⁹⁷ phosphorylation, whereas this inhibitor minimally affected eNOS Ser¹¹⁷⁹ phosphorylation stimulated by thrombin (Figure 5A) or the PAR2 agonist (data not shown). Y27632 also inhibited TFLLR (PAR1)-induced eNOS Thr⁴⁹⁷ phosphorylation (Figure 5B). However, Y27632 did not modulate cGMP production induced by thrombin in BAECs (Figure S2B).

Figure 4.

Requirement of G_12/13 for eNOS Thr⁴⁹⁷ phosphorylation stimulated by PAR1 in BAECs. The cells infected with adenovirus encoding p115RGS, a G_12/13 inhibitor, or control vector were stimulated with thrombin (10 U/mL) for 1 min or a PAR1 agonist TFLLR (50 μmol/L) for 2 min, and phosphorylation of eNOS and a ROCK substrate, MYPT, was evaluated. The bar graphs show densitometric analyses of the phosphorylation of eNOS and MYPT (mean ± SEM, *p < 0.05 compared with the basal control, †p < 0.05 compared with the stimulated control, bands scanned are n=3, each).

Figure 5.

Requirement of a Y27632-sensitive kinase for eNOS Thr⁴⁹⁷ phosphorylation stimulated by PAR1 in BAECs. In A and B, the cells were pretreated with Y27632 (10 μmol/L) for 30 min and stimulated with thrombin (10 U/mL) for 1 min (A) or a PAR1 agonist TFLLR (50 μmol/L) for 2 min (B). Phosphorylation of eNOS was evaluated. The bar graphs show densitometric analyses of the phosphorylation of eNOS at each site (mean ± SEM, *p < 0.05 compared with the stimulated control, bands scanned are n=3, each).

The specificity of Y27632 as a ROCK inhibitor has been questioned ¹⁶. H-1152, which has a better selectivity for ROCK than Y27632 ¹⁷, inhibited TFLLR-induced MYPT1 Thr⁸⁵³ phosphorylation but not eNOS Thr⁴⁹⁷ phosphorylation (Figure S3A). Also, dnRho markedly inhibited TFLLR-induced MYPT Thr⁸⁵³ phosphorylation but not eNOS Thr⁴⁹⁷ phosphorylation (Figure S3B). In addition, PKC did not contribute to the PAR1-induced eNOS Thr⁴⁹⁷ phosphorylation because it was insensitive to the PKC inhibitor GF109203X, whereas this inhibitor blocked the phorbol ester-induced eNOS Thr⁴⁹⁷ phosphorylation in BAECs (Figure S3C). Taken together, these data suggest that PAR1-induced phosphorylation of eNOS at Thr⁴⁹⁷ requires G_12/13 and subsequent activation of a Y27632-sensitive eNOS Thr⁴⁹⁷ kinase that is distinct from ROCK or PKC.

Discussion

The major findings of the present study are 1) stimulation of PAR2 in BAECs results in eNOS Ser¹¹⁷⁹ phosphorylation and subsequent NO production, 2) whereas stimulation of PAR1 solely stimulates eNOS Thr⁴⁹⁷ phosphorylation and does not lead to NO production, and 3) the downstream mechanisms utilized involve G_q for Ser¹¹⁷⁹ phosphorylation by PAR2, and G_12/13 and a previously un-identified Y27632-sensitive eNOS kinase for Thr⁴⁹⁷ phosphorylation by PAR1 (Figure 6). These data suggest distinct physiological and pathophysiological roles of PAR1 and PAR2 in regulating eNOS activity, representing novel mechanisms of PAR signal transduction in endothelial cells.

Figure 6.

Proposed signal transduction cascade of PAR1 and PAR2 in mediating the regulation of eNOS phosphoryation in endothelial cells. Note that upon thrombin stimulation of PAR1, there may be a competition between G_q and G_12/13, which favors G_q activation upon inhibition of G_12/13 and vise versa. In addition, G_12/13 inhibition might unmask the PAR1 coupling to eNOS Ser¹¹⁷⁹ phosphorylation via G_q.

In line with a theory of PAR2 transactivation by PAR1 proposed in HUVECs ³, our data suggest that both PAR1 and PAR2 mediate eNOS regulation by thrombin in BAECs. It is likely that thrombin at low concentration induces Thr⁴⁹⁷ phosphorylation through its high affinity receptor, PAR1, whereas at a higher concentration, thrombin further transactivates PAR2 through the PAR1 tethered ligand leading to Ser¹¹⁷⁹ phosphorylation and subsequent eNOS activation. This is further supported by the findings with a PAR1 tethered peptide used in the present study. Although a marked difference of G protein coupling affinity of PAR1 upon stimulation with thrombin or an agonistic peptide (the peptide prefers G_q coupling more than G_12/13) has been demonstrated ¹⁵, it appears not to be applicable to our findings. This is because no difference was observed between thrombin- or TFLLR- induced elevation of intracellular Ca²⁺ (G_q-dependent signal) or stimulation of MYPT phosphorylation (G_12/13-dependent signal) in BAECs.

We have observed that the G_q inhibitor YM-254890 enhanced thrombin-induced eNOS phosphorylation at Thr⁴⁹⁷. This is most likely due to competition of PAR1 coupling between G_q and G_12/13, which leads to the enhanced G_12/13 signal transduction upon inhibition of G_q. The presence of such competition is further supported by the enhanced Ser¹¹⁷⁹ phosphorylation by TFLLR or thrombin when G_12/13 activities were inhibited with p115RGS. Thus, G_12/13 inhibition might unmask the PAR1 coupling to eNOS Ser¹¹⁷⁹ phosphorylation via G_q as illustrated in Figure 6. However, it remains unclear why the phenomenon was not observed upon TFLLR stimulation with the G_q inhibitor. It is possible that TFLLR binding to the PAR1 may not be sufficient enough to change the competition which favors G_12/13 activation upon inhibition of G_q. In addition, TFLLR may not fully mimic the conformational change of the receptor induced by thrombin which causes cleavage of the receptor.

In the present study, we found that G_q is critical in PAR2-induced eNOS Ser¹¹⁷⁹ phosphorylation and subsequent enzymatic activation, as was recently shown in BAECs stimulated with thrombin or angiotensin II ^{8, 9}. Although thrombin-induced NO production has been shown to be markedly inhibited by Ca²⁺ chelators ⁹, the G_q-mediated Ca²⁺/CaM-dependent eNOS activation mechanism alone may be insufficient to stimulate NO production in BAECs. In this regard, we have previously demonstrated that thrombin-induced eNOS Ser¹¹⁷⁹ phosphorylation is mediated through a non-Akt kinase acting downstream of Ca²⁺, and that the Ser¹¹⁷⁹ phosphorylation is functionally indispensable for NO production stimulated by thrombin ⁹.

We have demonstrated the specificity of the G_q inhibitor, YM-254890, at the concentrations of 1-10 μmol/L in COS7 cells and vascular smooth muscle cells ^{13, 14}. Recently, the specificity was also verified in BAECs pretreated 30 min with 30 nmol/L of YM-254890. This concentration appeared to be sufficient to inhibit intracellular Ca²⁺ elevation by bradykinin. It also inhibited NO production induced by thrombin but not by ionomycin in BAECs ¹⁸. However, the concentrations greater than 50 nmol/L were required for the inhibition of eNOS Ser1179 phosphorylation by thrombin (Figure S1A) or ERK phosphorylation by angiotensin II ¹⁴, which is most likely due to the shorter treatment timw of 10 min.

Thors et al. proposed that Ser¹¹⁷⁹ phosphorylation of eNOS stimulated by thrombin is mediated through the Ca²⁺-dependent activation of an eNOS Ser¹¹⁷⁹-kinase, AMPK, by using a nonselective AMPK inhibitor in HUVECs ¹⁹. However, adenovirus transduction of dnAMPK did not prevent thrombin-induced eNOS Ser¹¹⁷⁹ phosphorylation in HUVECs ²⁰. In addition, the species differences in eNOS regulation by PARs ¹⁸ and the involvement of thrombomodulin in thrombin-induced eNOS phosphorylation have been reported ²¹. Therefore, further investigation is needed to identify the Ser¹¹⁷⁹ kinase as well as its exact upstream signal transduction utilized by PAR2.

Little was known about the regulation of eNOS Thr⁴⁹⁷ by GPCRs. We found that a Rho-kinase/ROCK inhibitor, Y27632, inhibited PAR1-induced eNOS Thr⁴⁹⁷ phosphorylation. Rho has been reported to inhibit NO production in arteries ²² and to inhibit eNOS activation through inhibition of Akt in endothelial cells ²³. In this regard, a recent study proposed ROCK as the eNOS Thr⁴⁹⁷ kinase activated by thrombin ²⁴. Although this study demonstrated that ROCK was able to phosphorylate eNOS in vitro, only Y27632 was used to block the phosphorylation in vivo. Our data using dnRho and a more selective ROCK inhibitor rather suggest the presence of a Y27632-sensitive novel eNOS Thr⁴⁹⁷ kinase distinct from the Rho/ROCK.

A negative regulatory role of the eNOS Thr⁴⁹⁷ phosphorylation has been reported ²⁵, however, we could not observe enhancement of cGMP production upon inhibition of the phosphorylation with Y27632 in the present study. A mutational experiment of eNOS Thr⁴⁹⁷ to mimic constitutive phosphorylation (Asp⁴⁹⁵ in human eNOS) with in vitro measurement of the enzymatic activity suggests that eNOS Thr⁴⁹⁷ phosphorylation reduces Ca²⁺ sensitivity of the enzyme ²⁵. However, expression and stimulation of an eNOS Thr⁴⁹⁷ phosphorylation mutant to mimic de-phosphorylation (Ala⁴⁹⁷) expressed in COS7 cells did not enhance NO production over that of wild type, whereas a mutation in Asp⁴⁹⁷ inhibited NO production ²⁶. Taken together with our Ca²⁺ stimulation data by PAR agonists, it is likely that eNOS phosphorylation at Thr⁴⁹⁷ blocks eNOS activity against Ca²⁺/CaM but is insufficient to block the activity with a concurrent phosphorylation of Ser¹¹⁷⁹.

Although we have observed similar regulation of eNOS by PARs in BAECs and HUVECs in the present study, the expression ratio of PARs may be different in endothelial cells from distinct species and/or vascular beds as may be that of G proteins as well. Moreover, expression of PARs in endothelial cells is under the regulation of distinct extracellular conditions such as inflammation ²⁷. PAR1 mRNA was upregulated in rat aorta associated with angiotensin II-induced hypertension ²⁸. However, exact roles of PARs in vascular tonus regulation still remain unclear. Therefore, additional experiments in various in vivo settings will be necessary to better generalize our findings in certain vascular pathophysiology such as in hypertension.

Perspectives

Selective manipulation of G_q or G_12/13 in vascular smooth muscle cells revealed the importance of these G proteins in the etiology of high blood pressure ²⁹. The close link between endothelial G₁₃ and PAR1 has been demonstrated ³⁰. Moreover, positive regulation of eNOS expression by G₁₂ has been reported ³¹. Further detailed research specifically on eNOS phosphorylation regulation by PARs will shed light on critical mechanisms by which multiple GPCRs expressed in the endothelium potentially regulate endothelial dysfunction associated with cardiovascular diseases.